From 4c7f40287aa5eb6ea8f3a6c3225ab5ebc6a45415 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Mon, 31 Aug 2020 13:22:20 +0800 Subject: [PATCH 01/13] net/mlx5: E-switch, Move vport table functions to a new file Currently, the vport table functions are in common eswitch offload file. This file is too big. Move the vport table create, delete and lookup functions to a separate file. Put the file in esw directory. Pre-step for generalizing its functionality for serving both the mirroring and the sample features. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/Makefile | 2 +- .../mellanox/mlx5/core/esw/vporttbl.c | 136 +++++++++++ .../net/ethernet/mellanox/mlx5/core/eswitch.h | 12 +- .../mellanox/mlx5/core/eswitch_offloads.c | 215 +++--------------- 4 files changed, 183 insertions(+), 182 deletions(-) create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index 9cf7de72df52..4c86b68ad26c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -53,7 +53,7 @@ mlx5_core-$(CONFIG_MLX5_ESWITCH) += eswitch.o eswitch_offloads.o eswitch_offlo mlx5_core-$(CONFIG_MLX5_ESWITCH) += esw/acl/helper.o \ esw/acl/egress_lgcy.o esw/acl/egress_ofld.o \ esw/acl/ingress_lgcy.o esw/acl/ingress_ofld.o \ - esw/devlink_port.o + esw/devlink_port.o esw/vporttbl.o mlx5_core-$(CONFIG_MLX5_MPFS) += lib/mpfs.o mlx5_core-$(CONFIG_VXLAN) += lib/vxlan.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c new file mode 100644 index 000000000000..8219c5d50db0 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c @@ -0,0 +1,136 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +// Copyright (c) 2021 Mellanox Technologies. + +#include "eswitch.h" + +#define MLX5_ESW_VPORT_TABLE_SIZE 128 +#define MLX5_ESW_VPORT_TBL_NUM_GROUPS 4 + +/* This struct is used as a key to the hash table and we need it to be packed + * so hash result is consistent + */ +struct mlx5_vport_key { + u32 chain; + u16 prio; + u16 vport; + u16 vhca_id; +} __packed; + +struct mlx5_vport_table { + struct hlist_node hlist; + struct mlx5_flow_table *fdb; + u32 num_rules; + struct mlx5_vport_key key; +}; + +static struct mlx5_flow_table * +esw_vport_tbl_create(struct mlx5_eswitch *esw, struct mlx5_flow_namespace *ns) +{ + struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5_flow_table *fdb; + + ft_attr.autogroup.max_num_groups = MLX5_ESW_VPORT_TBL_NUM_GROUPS; + ft_attr.max_fte = MLX5_ESW_VPORT_TABLE_SIZE; + ft_attr.prio = FDB_PER_VPORT; + fdb = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); + if (IS_ERR(fdb)) { + esw_warn(esw->dev, "Failed to create per vport FDB Table err %ld\n", + PTR_ERR(fdb)); + } + + return fdb; +} + +static u32 flow_attr_to_vport_key(struct mlx5_eswitch *esw, + struct mlx5_vport_tbl_attr *attr, + struct mlx5_vport_key *key) +{ + key->vport = attr->vport; + key->chain = attr->chain; + key->prio = attr->prio; + key->vhca_id = MLX5_CAP_GEN(esw->dev, vhca_id); + return jhash(key, sizeof(*key), 0); +} + +/* caller must hold vports.lock */ +static struct mlx5_vport_table * +esw_vport_tbl_lookup(struct mlx5_eswitch *esw, struct mlx5_vport_key *skey, u32 key) +{ + struct mlx5_vport_table *e; + + hash_for_each_possible(esw->fdb_table.offloads.vports.table, e, hlist, key) + if (!memcmp(&e->key, skey, sizeof(*skey))) + return e; + + return NULL; +} + +struct mlx5_flow_table * +esw_vport_tbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) +{ + struct mlx5_core_dev *dev = esw->dev; + struct mlx5_flow_namespace *ns; + struct mlx5_flow_table *fdb; + struct mlx5_vport_table *e; + struct mlx5_vport_key skey; + u32 hkey; + + mutex_lock(&esw->fdb_table.offloads.vports.lock); + hkey = flow_attr_to_vport_key(esw, attr, &skey); + e = esw_vport_tbl_lookup(esw, &skey, hkey); + if (e) { + e->num_rules++; + goto out; + } + + e = kzalloc(sizeof(*e), GFP_KERNEL); + if (!e) { + fdb = ERR_PTR(-ENOMEM); + goto err_alloc; + } + + ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_FDB); + if (!ns) { + esw_warn(dev, "Failed to get FDB namespace\n"); + fdb = ERR_PTR(-ENOENT); + goto err_ns; + } + + fdb = esw_vport_tbl_create(esw, ns); + if (IS_ERR(fdb)) + goto err_ns; + + e->fdb = fdb; + e->num_rules = 1; + e->key = skey; + hash_add(esw->fdb_table.offloads.vports.table, &e->hlist, hkey); +out: + mutex_unlock(&esw->fdb_table.offloads.vports.lock); + return e->fdb; + +err_ns: + kfree(e); +err_alloc: + mutex_unlock(&esw->fdb_table.offloads.vports.lock); + return fdb; +} + +void +esw_vport_tbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) +{ + struct mlx5_vport_table *e; + struct mlx5_vport_key key; + u32 hkey; + + mutex_lock(&esw->fdb_table.offloads.vports.lock); + hkey = flow_attr_to_vport_key(esw, attr, &key); + e = esw_vport_tbl_lookup(esw, &key, hkey); + if (!e || --e->num_rules) + goto out; + + hash_del(&e->hlist); + mlx5_destroy_flow_table(e->fdb); + kfree(e); +out: + mutex_unlock(&esw->fdb_table.offloads.vports.lock); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 64db903068c1..70eeb8d1ae03 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -713,8 +713,16 @@ void esw_vport_destroy_offloads_acl_tables(struct mlx5_eswitch *esw, struct mlx5_vport *vport); -int mlx5_esw_vport_tbl_get(struct mlx5_eswitch *esw); -void mlx5_esw_vport_tbl_put(struct mlx5_eswitch *esw); +struct mlx5_vport_tbl_attr { + u16 chain; + u16 prio; + u16 vport; +}; + +struct mlx5_flow_table * +esw_vport_tbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr); +void +esw_vport_tbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr); struct mlx5_flow_handle * esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index d5de6bf622ce..3caf6c0b3296 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -54,185 +54,6 @@ #define MLX5_ESW_MISS_FLOWS (2) #define UPLINK_REP_INDEX 0 -/* Per vport tables */ - -#define MLX5_ESW_VPORT_TABLE_SIZE 128 - -/* This struct is used as a key to the hash table and we need it to be packed - * so hash result is consistent - */ -struct mlx5_vport_key { - u32 chain; - u16 prio; - u16 vport; - u16 vhca_id; -} __packed; - -struct mlx5_vport_tbl_attr { - u16 chain; - u16 prio; - u16 vport; -}; - -struct mlx5_vport_table { - struct hlist_node hlist; - struct mlx5_flow_table *fdb; - u32 num_rules; - struct mlx5_vport_key key; -}; - -#define MLX5_ESW_VPORT_TBL_NUM_GROUPS 4 - -static struct mlx5_flow_table * -esw_vport_tbl_create(struct mlx5_eswitch *esw, struct mlx5_flow_namespace *ns) -{ - struct mlx5_flow_table_attr ft_attr = {}; - struct mlx5_flow_table *fdb; - - ft_attr.autogroup.max_num_groups = MLX5_ESW_VPORT_TBL_NUM_GROUPS; - ft_attr.max_fte = MLX5_ESW_VPORT_TABLE_SIZE; - ft_attr.prio = FDB_PER_VPORT; - fdb = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); - if (IS_ERR(fdb)) { - esw_warn(esw->dev, "Failed to create per vport FDB Table err %ld\n", - PTR_ERR(fdb)); - } - - return fdb; -} - -static u32 flow_attr_to_vport_key(struct mlx5_eswitch *esw, - struct mlx5_vport_tbl_attr *attr, - struct mlx5_vport_key *key) -{ - key->vport = attr->vport; - key->chain = attr->chain; - key->prio = attr->prio; - key->vhca_id = MLX5_CAP_GEN(esw->dev, vhca_id); - return jhash(key, sizeof(*key), 0); -} - -/* caller must hold vports.lock */ -static struct mlx5_vport_table * -esw_vport_tbl_lookup(struct mlx5_eswitch *esw, struct mlx5_vport_key *skey, u32 key) -{ - struct mlx5_vport_table *e; - - hash_for_each_possible(esw->fdb_table.offloads.vports.table, e, hlist, key) - if (!memcmp(&e->key, skey, sizeof(*skey))) - return e; - - return NULL; -} - -static void -esw_vport_tbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) -{ - struct mlx5_vport_table *e; - struct mlx5_vport_key key; - u32 hkey; - - mutex_lock(&esw->fdb_table.offloads.vports.lock); - hkey = flow_attr_to_vport_key(esw, attr, &key); - e = esw_vport_tbl_lookup(esw, &key, hkey); - if (!e || --e->num_rules) - goto out; - - hash_del(&e->hlist); - mlx5_destroy_flow_table(e->fdb); - kfree(e); -out: - mutex_unlock(&esw->fdb_table.offloads.vports.lock); -} - -static struct mlx5_flow_table * -esw_vport_tbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) -{ - struct mlx5_core_dev *dev = esw->dev; - struct mlx5_flow_namespace *ns; - struct mlx5_flow_table *fdb; - struct mlx5_vport_table *e; - struct mlx5_vport_key skey; - u32 hkey; - - mutex_lock(&esw->fdb_table.offloads.vports.lock); - hkey = flow_attr_to_vport_key(esw, attr, &skey); - e = esw_vport_tbl_lookup(esw, &skey, hkey); - if (e) { - e->num_rules++; - goto out; - } - - e = kzalloc(sizeof(*e), GFP_KERNEL); - if (!e) { - fdb = ERR_PTR(-ENOMEM); - goto err_alloc; - } - - ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_FDB); - if (!ns) { - esw_warn(dev, "Failed to get FDB namespace\n"); - fdb = ERR_PTR(-ENOENT); - goto err_ns; - } - - fdb = esw_vport_tbl_create(esw, ns); - if (IS_ERR(fdb)) - goto err_ns; - - e->fdb = fdb; - e->num_rules = 1; - e->key = skey; - hash_add(esw->fdb_table.offloads.vports.table, &e->hlist, hkey); -out: - mutex_unlock(&esw->fdb_table.offloads.vports.lock); - return e->fdb; - -err_ns: - kfree(e); -err_alloc: - mutex_unlock(&esw->fdb_table.offloads.vports.lock); - return fdb; -} - -int mlx5_esw_vport_tbl_get(struct mlx5_eswitch *esw) -{ - struct mlx5_vport_tbl_attr attr; - struct mlx5_flow_table *fdb; - struct mlx5_vport *vport; - int i; - - attr.chain = 0; - attr.prio = 1; - mlx5_esw_for_all_vports(esw, i, vport) { - attr.vport = vport->vport; - fdb = esw_vport_tbl_get(esw, &attr); - if (IS_ERR(fdb)) - goto out; - } - return 0; - -out: - mlx5_esw_vport_tbl_put(esw); - return PTR_ERR(fdb); -} - -void mlx5_esw_vport_tbl_put(struct mlx5_eswitch *esw) -{ - struct mlx5_vport_tbl_attr attr; - struct mlx5_vport *vport; - int i; - - attr.chain = 0; - attr.prio = 1; - mlx5_esw_for_all_vports(esw, i, vport) { - attr.vport = vport->vport; - esw_vport_tbl_put(esw, &attr); - } -} - -/* End: Per vport tables */ - static struct mlx5_eswitch_rep *mlx5_eswitch_get_rep(struct mlx5_eswitch *esw, u16 vport_num) { @@ -1514,6 +1335,42 @@ static void esw_set_flow_group_source_port(struct mlx5_eswitch *esw, } #if IS_ENABLED(CONFIG_MLX5_CLS_ACT) +static void mlx5_esw_vport_tbl_put(struct mlx5_eswitch *esw) +{ + struct mlx5_vport_tbl_attr attr; + struct mlx5_vport *vport; + int i; + + attr.chain = 0; + attr.prio = 1; + mlx5_esw_for_all_vports(esw, i, vport) { + attr.vport = vport->vport; + esw_vport_tbl_put(esw, &attr); + } +} + +static int mlx5_esw_vport_tbl_get(struct mlx5_eswitch *esw) +{ + struct mlx5_vport_tbl_attr attr; + struct mlx5_flow_table *fdb; + struct mlx5_vport *vport; + int i; + + attr.chain = 0; + attr.prio = 1; + mlx5_esw_for_all_vports(esw, i, vport) { + attr.vport = vport->vport; + fdb = esw_vport_tbl_get(esw, &attr); + if (IS_ERR(fdb)) + goto out; + } + return 0; + +out: + mlx5_esw_vport_tbl_put(esw); + return PTR_ERR(fdb); +} + #define fdb_modify_header_fwd_to_table_supported(esw) \ (MLX5_CAP_ESW_FLOWTABLE((esw)->dev, fdb_modify_header_fwd_to_table)) static void esw_init_chains_offload_flags(struct mlx5_eswitch *esw, u32 *flags) From 0a9e2307870bdb627421b4b0bbac04b1ab8e1688 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Thu, 14 Jan 2021 15:12:36 +0800 Subject: [PATCH 02/13] net/mlx5: E-switch, Rename functions to follow naming convention. Public api starts with mlx5 and remove mlx5 for non-public api. Signed-off-by: Chris Mi Signed-off-by: Saeed Mahameed --- .../mellanox/mlx5/core/esw/vporttbl.c | 4 +-- .../net/ethernet/mellanox/mlx5/core/eswitch.h | 4 +-- .../mellanox/mlx5/core/eswitch_offloads.c | 26 +++++++++---------- 3 files changed, 17 insertions(+), 17 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c index 8219c5d50db0..6c4246181615 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c @@ -66,7 +66,7 @@ esw_vport_tbl_lookup(struct mlx5_eswitch *esw, struct mlx5_vport_key *skey, u32 } struct mlx5_flow_table * -esw_vport_tbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) +mlx5_esw_vporttbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) { struct mlx5_core_dev *dev = esw->dev; struct mlx5_flow_namespace *ns; @@ -116,7 +116,7 @@ err_alloc: } void -esw_vport_tbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) +mlx5_esw_vporttbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr) { struct mlx5_vport_table *e; struct mlx5_vport_key key; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 70eeb8d1ae03..b7d1f8854ef4 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -720,9 +720,9 @@ struct mlx5_vport_tbl_attr { }; struct mlx5_flow_table * -esw_vport_tbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr); +mlx5_esw_vporttbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr); void -esw_vport_tbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr); +mlx5_esw_vporttbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr); struct mlx5_flow_handle * esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 3caf6c0b3296..63e22e9e5ad1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -483,7 +483,7 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, fwd_attr.prio = attr->prio; fwd_attr.vport = esw_attr->in_rep->vport; - fdb = esw_vport_tbl_get(esw, &fwd_attr); + fdb = mlx5_esw_vporttbl_get(esw, &fwd_attr); } else { if (attr->chain || attr->prio) fdb = mlx5_chains_get_table(chains, attr->chain, @@ -515,7 +515,7 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, err_add_rule: if (split) - esw_vport_tbl_put(esw, &fwd_attr); + mlx5_esw_vporttbl_put(esw, &fwd_attr); else if (attr->chain || attr->prio) mlx5_chains_put_table(chains, attr->chain, attr->prio, 0); err_esw_get: @@ -548,7 +548,7 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, fwd_attr.chain = attr->chain; fwd_attr.prio = attr->prio; fwd_attr.vport = esw_attr->in_rep->vport; - fwd_fdb = esw_vport_tbl_get(esw, &fwd_attr); + fwd_fdb = mlx5_esw_vporttbl_get(esw, &fwd_attr); if (IS_ERR(fwd_fdb)) { rule = ERR_CAST(fwd_fdb); goto err_get_fwd; @@ -593,7 +593,7 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, return rule; err_chain_src_rewrite: esw_put_dest_tables_loop(esw, attr, 0, i); - esw_vport_tbl_put(esw, &fwd_attr); + mlx5_esw_vporttbl_put(esw, &fwd_attr); err_get_fwd: mlx5_chains_put_table(chains, attr->chain, attr->prio, 0); err_get_fast: @@ -631,12 +631,12 @@ __mlx5_eswitch_del_rule(struct mlx5_eswitch *esw, } if (fwd_rule) { - esw_vport_tbl_put(esw, &fwd_attr); + mlx5_esw_vporttbl_put(esw, &fwd_attr); mlx5_chains_put_table(chains, attr->chain, attr->prio, 0); esw_put_dest_tables_loop(esw, attr, 0, esw_attr->split_count); } else { if (split) - esw_vport_tbl_put(esw, &fwd_attr); + mlx5_esw_vporttbl_put(esw, &fwd_attr); else if (attr->chain || attr->prio) mlx5_chains_put_table(chains, attr->chain, attr->prio, 0); esw_cleanup_dests(esw, attr); @@ -1335,7 +1335,7 @@ static void esw_set_flow_group_source_port(struct mlx5_eswitch *esw, } #if IS_ENABLED(CONFIG_MLX5_CLS_ACT) -static void mlx5_esw_vport_tbl_put(struct mlx5_eswitch *esw) +static void esw_vport_tbl_put(struct mlx5_eswitch *esw) { struct mlx5_vport_tbl_attr attr; struct mlx5_vport *vport; @@ -1345,11 +1345,11 @@ static void mlx5_esw_vport_tbl_put(struct mlx5_eswitch *esw) attr.prio = 1; mlx5_esw_for_all_vports(esw, i, vport) { attr.vport = vport->vport; - esw_vport_tbl_put(esw, &attr); + mlx5_esw_vporttbl_put(esw, &attr); } } -static int mlx5_esw_vport_tbl_get(struct mlx5_eswitch *esw) +static int esw_vport_tbl_get(struct mlx5_eswitch *esw) { struct mlx5_vport_tbl_attr attr; struct mlx5_flow_table *fdb; @@ -1360,14 +1360,14 @@ static int mlx5_esw_vport_tbl_get(struct mlx5_eswitch *esw) attr.prio = 1; mlx5_esw_for_all_vports(esw, i, vport) { attr.vport = vport->vport; - fdb = esw_vport_tbl_get(esw, &attr); + fdb = mlx5_esw_vporttbl_get(esw, &attr); if (IS_ERR(fdb)) goto out; } return 0; out: - mlx5_esw_vport_tbl_put(esw); + esw_vport_tbl_put(esw); return PTR_ERR(fdb); } @@ -1448,7 +1448,7 @@ esw_chains_create(struct mlx5_eswitch *esw, struct mlx5_flow_table *miss_fdb) /* Open level 1 for split fdb rules now if prios isn't supported */ if (!mlx5_chains_prios_supported(chains)) { - err = mlx5_esw_vport_tbl_get(esw); + err = esw_vport_tbl_get(esw); if (err) goto level_1_err; } @@ -1472,7 +1472,7 @@ static void esw_chains_destroy(struct mlx5_eswitch *esw, struct mlx5_fs_chains *chains) { if (!mlx5_chains_prios_supported(chains)) - mlx5_esw_vport_tbl_put(esw); + esw_vport_tbl_put(esw); mlx5_chains_put_table(chains, 0, 1, 0); mlx5_chains_put_table(chains, mlx5_chains_get_nf_ft_chain(chains), 1, 0); mlx5_chains_destroy(chains); From c796bb7cd2307d5fb815e919b964bf2d65c95ec3 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Mon, 31 Aug 2020 13:23:25 +0800 Subject: [PATCH 03/13] net/mlx5: E-switch, Generalize per vport table API Currently, per vport table was used only for port mirroring actions. However, sample action will also require a per vport table instance. Generalize the vport table API to work with multiple namespaces where each namespace manages its own vport table instance. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/esw/vporttbl.c | 15 ++++++++------- drivers/net/ethernet/mellanox/mlx5/core/eswitch.h | 7 +++++++ .../mellanox/mlx5/core/eswitch_offloads.c | 14 ++++++++++++++ 3 files changed, 29 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c index 6c4246181615..abba1b801048 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c @@ -3,9 +3,6 @@ #include "eswitch.h" -#define MLX5_ESW_VPORT_TABLE_SIZE 128 -#define MLX5_ESW_VPORT_TBL_NUM_GROUPS 4 - /* This struct is used as a key to the hash table and we need it to be packed * so hash result is consistent */ @@ -14,6 +11,7 @@ struct mlx5_vport_key { u16 prio; u16 vport; u16 vhca_id; + const struct esw_vport_tbl_namespace *vport_ns; } __packed; struct mlx5_vport_table { @@ -24,14 +22,16 @@ struct mlx5_vport_table { }; static struct mlx5_flow_table * -esw_vport_tbl_create(struct mlx5_eswitch *esw, struct mlx5_flow_namespace *ns) +esw_vport_tbl_create(struct mlx5_eswitch *esw, struct mlx5_flow_namespace *ns, + const struct esw_vport_tbl_namespace *vport_ns) { struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_flow_table *fdb; - ft_attr.autogroup.max_num_groups = MLX5_ESW_VPORT_TBL_NUM_GROUPS; - ft_attr.max_fte = MLX5_ESW_VPORT_TABLE_SIZE; + ft_attr.autogroup.max_num_groups = vport_ns->max_num_groups; + ft_attr.max_fte = vport_ns->max_fte; ft_attr.prio = FDB_PER_VPORT; + ft_attr.flags = vport_ns->flags; fdb = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); if (IS_ERR(fdb)) { esw_warn(esw->dev, "Failed to create per vport FDB Table err %ld\n", @@ -49,6 +49,7 @@ static u32 flow_attr_to_vport_key(struct mlx5_eswitch *esw, key->chain = attr->chain; key->prio = attr->prio; key->vhca_id = MLX5_CAP_GEN(esw->dev, vhca_id); + key->vport_ns = attr->vport_ns; return jhash(key, sizeof(*key), 0); } @@ -96,7 +97,7 @@ mlx5_esw_vporttbl_get(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr goto err_ns; } - fdb = esw_vport_tbl_create(esw, ns); + fdb = esw_vport_tbl_create(esw, ns, attr->vport_ns); if (IS_ERR(fdb)) goto err_ns; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index b7d1f8854ef4..e0415676821a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -713,10 +713,17 @@ void esw_vport_destroy_offloads_acl_tables(struct mlx5_eswitch *esw, struct mlx5_vport *vport); +struct esw_vport_tbl_namespace { + int max_fte; + int max_num_groups; + u32 flags; +}; + struct mlx5_vport_tbl_attr { u16 chain; u16 prio; u16 vport; + const struct esw_vport_tbl_namespace *vport_ns; }; struct mlx5_flow_table * diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 63e22e9e5ad1..8ac4b60ea225 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -54,6 +54,15 @@ #define MLX5_ESW_MISS_FLOWS (2) #define UPLINK_REP_INDEX 0 +#define MLX5_ESW_VPORT_TBL_SIZE 128 +#define MLX5_ESW_VPORT_TBL_NUM_GROUPS 4 + +static const struct esw_vport_tbl_namespace mlx5_esw_vport_tbl_mirror_ns = { + .max_fte = MLX5_ESW_VPORT_TBL_SIZE, + .max_num_groups = MLX5_ESW_VPORT_TBL_NUM_GROUPS, + .flags = 0, +}; + static struct mlx5_eswitch_rep *mlx5_eswitch_get_rep(struct mlx5_eswitch *esw, u16 vport_num) { @@ -482,6 +491,7 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, fwd_attr.chain = attr->chain; fwd_attr.prio = attr->prio; fwd_attr.vport = esw_attr->in_rep->vport; + fwd_attr.vport_ns = &mlx5_esw_vport_tbl_mirror_ns; fdb = mlx5_esw_vporttbl_get(esw, &fwd_attr); } else { @@ -548,6 +558,7 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, fwd_attr.chain = attr->chain; fwd_attr.prio = attr->prio; fwd_attr.vport = esw_attr->in_rep->vport; + fwd_attr.vport_ns = &mlx5_esw_vport_tbl_mirror_ns; fwd_fdb = mlx5_esw_vporttbl_get(esw, &fwd_attr); if (IS_ERR(fwd_fdb)) { rule = ERR_CAST(fwd_fdb); @@ -628,6 +639,7 @@ __mlx5_eswitch_del_rule(struct mlx5_eswitch *esw, fwd_attr.chain = attr->chain; fwd_attr.prio = attr->prio; fwd_attr.vport = esw_attr->in_rep->vport; + fwd_attr.vport_ns = &mlx5_esw_vport_tbl_mirror_ns; } if (fwd_rule) { @@ -1345,6 +1357,7 @@ static void esw_vport_tbl_put(struct mlx5_eswitch *esw) attr.prio = 1; mlx5_esw_for_all_vports(esw, i, vport) { attr.vport = vport->vport; + attr.vport_ns = &mlx5_esw_vport_tbl_mirror_ns; mlx5_esw_vporttbl_put(esw, &attr); } } @@ -1360,6 +1373,7 @@ static int esw_vport_tbl_get(struct mlx5_eswitch *esw) attr.prio = 1; mlx5_esw_for_all_vports(esw, i, vport) { attr.vport = vport->vport; + attr.vport_ns = &mlx5_esw_vport_tbl_mirror_ns; fdb = mlx5_esw_vporttbl_get(esw, &attr); if (IS_ERR(fdb)) goto out; From c1904360dde8fb4f79bec4751c7540df10d57ec2 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Fri, 9 Oct 2020 11:06:33 +0800 Subject: [PATCH 04/13] net/mlx5: E-switch, Set per vport table default group number Different per voprt table is created using a different per vport table namespace. Because we can't use variable to set the namespace member value. If max group number is 0 in the namespace, use the eswitch default max group number. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c index abba1b801048..9e72118f2e4c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/vporttbl.c @@ -28,7 +28,10 @@ esw_vport_tbl_create(struct mlx5_eswitch *esw, struct mlx5_flow_namespace *ns, struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_flow_table *fdb; - ft_attr.autogroup.max_num_groups = vport_ns->max_num_groups; + if (vport_ns->max_num_groups) + ft_attr.autogroup.max_num_groups = vport_ns->max_num_groups; + else + ft_attr.autogroup.max_num_groups = esw->params.large_group_num; ft_attr.max_fte = vport_ns->max_fte; ft_attr.prio = FDB_PER_VPORT; ft_attr.flags = vport_ns->flags; From a91d98a0a2b8e4c433b7341708f7d706e0cf1c8e Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Thu, 10 Sep 2020 15:28:02 +0800 Subject: [PATCH 05/13] net/mlx5: Map register values to restore objects Currently reg_c0 lower 16 bits and reg_b are used to store the chain id that missed in FDB and NIC tables accordingly. However, the registers' values may index a restore object, rather than a single u32 value. Different object types can be used to restore mutually exclusive contexts such as chain id and sample group id. Use the mapping object to associate an index with a restore object as a prestep for supporting additional restore types. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/en/rep/tc.c | 42 +++++++++++-------- .../net/ethernet/mellanox/mlx5/core/en_tc.c | 9 +++- .../net/ethernet/mellanox/mlx5/core/eswitch.h | 13 +++++- .../mellanox/mlx5/core/eswitch_offloads.c | 14 ++----- .../mellanox/mlx5/core/lib/fs_chains.c | 20 +++++---- .../mellanox/mlx5/core/lib/fs_chains.h | 3 +- include/linux/mlx5/eswitch.h | 9 ++-- 7 files changed, 65 insertions(+), 45 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c index 11a44d30adc7..dde83cba85c3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c @@ -618,9 +618,10 @@ bool mlx5e_rep_tc_update_skb(struct mlx5_cqe64 *cqe, struct mlx5e_tc_update_priv *tc_priv) { #if IS_ENABLED(CONFIG_NET_TC_SKB_EXT) - u32 chain = 0, reg_c0, reg_c1, tunnel_id, zone_restore_id; + u32 reg_c0, reg_c1, tunnel_id, zone_restore_id; struct mlx5_rep_uplink_priv *uplink_priv; struct mlx5e_rep_priv *uplink_rpriv; + struct mlx5_mapped_obj mapped_obj; struct tc_skb_ext *tc_skb_ext; struct mlx5_eswitch *esw; struct mlx5e_priv *priv; @@ -640,30 +641,35 @@ bool mlx5e_rep_tc_update_skb(struct mlx5_cqe64 *cqe, priv = netdev_priv(skb->dev); esw = priv->mdev->priv.eswitch; - err = mlx5_get_chain_for_tag(esw_chains(esw), reg_c0, &chain); + err = mlx5_get_mapped_object(esw_chains(esw), reg_c0, &mapped_obj); if (err) { netdev_dbg(priv->netdev, - "Couldn't find chain for chain tag: %d, err: %d\n", + "Couldn't find mapped object for reg_c0: %d, err: %d\n", reg_c0, err); return false; } - if (chain) { - tc_skb_ext = skb_ext_add(skb, TC_SKB_EXT); - if (!tc_skb_ext) { - WARN_ON(1); - return false; + if (mapped_obj.type == MLX5_MAPPED_OBJ_CHAIN) { + if (mapped_obj.chain) { + tc_skb_ext = skb_ext_add(skb, TC_SKB_EXT); + if (!tc_skb_ext) { + WARN_ON(1); + return false; + } + + tc_skb_ext->chain = mapped_obj.chain; + + zone_restore_id = reg_c1 & ESW_ZONE_ID_MASK; + + uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); + uplink_priv = &uplink_rpriv->uplink_priv; + if (!mlx5e_tc_ct_restore_flow(uplink_priv->ct_priv, skb, + zone_restore_id)) + return false; } - - tc_skb_ext->chain = chain; - - zone_restore_id = reg_c1 & ESW_ZONE_ID_MASK; - - uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); - uplink_priv = &uplink_rpriv->uplink_priv; - if (!mlx5e_tc_ct_restore_flow(uplink_priv->ct_priv, skb, - zone_restore_id)) - return false; + } else { + netdev_dbg(priv->netdev, "Invalid mapped object type: %d\n", mapped_obj.type); + return false; } tunnel_id = reg_c1 >> ESW_TUN_OFFSET; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index bb1e0d442b5c..9b5607ddb9a2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -4973,6 +4973,7 @@ bool mlx5e_tc_update_skb(struct mlx5_cqe64 *cqe, u32 chain = 0, chain_tag, reg_b, zone_restore_id; struct mlx5e_priv *priv = netdev_priv(skb->dev); struct mlx5e_tc_table *tc = &priv->fs.tc; + struct mlx5_mapped_obj mapped_obj; struct tc_skb_ext *tc_skb_ext; int err; @@ -4980,7 +4981,7 @@ bool mlx5e_tc_update_skb(struct mlx5_cqe64 *cqe, chain_tag = reg_b & MLX5E_TC_TABLE_CHAIN_TAG_MASK; - err = mlx5_get_chain_for_tag(nic_chains(priv), chain_tag, &chain); + err = mlx5_get_mapped_object(nic_chains(priv), chain_tag, &mapped_obj); if (err) { netdev_dbg(priv->netdev, "Couldn't find chain for chain tag: %d, err: %d\n", @@ -4988,7 +4989,8 @@ bool mlx5e_tc_update_skb(struct mlx5_cqe64 *cqe, return false; } - if (chain) { + if (mapped_obj.type == MLX5_MAPPED_OBJ_CHAIN) { + chain = mapped_obj.chain; tc_skb_ext = skb_ext_add(skb, TC_SKB_EXT); if (WARN_ON(!tc_skb_ext)) return false; @@ -5001,6 +5003,9 @@ bool mlx5e_tc_update_skb(struct mlx5_cqe64 *cqe, if (!mlx5e_tc_ct_restore_flow(tc->ct, skb, zone_restore_id)) return false; + } else { + netdev_dbg(priv->netdev, "Invalid mapped object type: %d\n", mapped_obj.type); + return false; } #endif /* CONFIG_NET_TC_SKB_EXT */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index e0415676821a..c5b35e7f8aed 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -47,6 +47,17 @@ #include "sf/sf.h" #include "en/tc_ct.h" +enum mlx5_mapped_obj_type { + MLX5_MAPPED_OBJ_CHAIN, +}; + +struct mlx5_mapped_obj { + enum mlx5_mapped_obj_type type; + union { + u32 chain; + }; +}; + #ifdef CONFIG_MLX5_ESWITCH #define ESW_OFFLOADS_DEFAULT_NUM_GROUPS 15 @@ -733,8 +744,6 @@ mlx5_esw_vporttbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr struct mlx5_flow_handle * esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag); -u32 -esw_get_max_restore_tag(struct mlx5_eswitch *esw); int esw_offloads_load_rep(struct mlx5_eswitch *esw, u16 vport_num); void esw_offloads_unload_rep(struct mlx5_eswitch *esw, u16 vport_num); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 8ac4b60ea225..117d9fa93ff5 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -1286,7 +1286,7 @@ esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag) misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters_2); MLX5_SET(fte_match_set_misc2, misc, metadata_reg_c_0, - ESW_CHAIN_TAG_METADATA_MASK); + ESW_REG_C0_USER_DATA_METADATA_MASK); misc = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters_2); MLX5_SET(fte_match_set_misc2, misc, metadata_reg_c_0, tag); @@ -1312,12 +1312,6 @@ esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag) return flow_rule; } -u32 -esw_get_max_restore_tag(struct mlx5_eswitch *esw) -{ - return ESW_CHAIN_TAG_METADATA_MASK; -} - #define MAX_PF_SQ 256 #define MAX_SQ_NVPORTS 32 @@ -1434,7 +1428,7 @@ esw_chains_create(struct mlx5_eswitch *esw, struct mlx5_flow_table *miss_fdb) attr.max_ft_sz = fdb_max; attr.max_grp_num = esw->params.large_group_num; attr.default_ft = miss_fdb; - attr.max_restore_tag = esw_get_max_restore_tag(esw); + attr.max_restore_tag = ESW_REG_C0_USER_DATA_METADATA_MASK; chains = mlx5_chains_create(dev, &attr); if (IS_ERR(chains)) { @@ -1928,7 +1922,7 @@ static int esw_create_restore_table(struct mlx5_eswitch *esw) goto out_free; } - ft_attr.max_fte = 1 << ESW_CHAIN_TAG_METADATA_BITS; + ft_attr.max_fte = 1 << ESW_REG_C0_USER_DATA_METADATA_BITS; ft = mlx5_create_flow_table(ns, &ft_attr); if (IS_ERR(ft)) { err = PTR_ERR(ft); @@ -1943,7 +1937,7 @@ static int esw_create_restore_table(struct mlx5_eswitch *esw) misc_parameters_2); MLX5_SET(fte_match_set_misc2, misc, metadata_reg_c_0, - ESW_CHAIN_TAG_METADATA_MASK); + ESW_REG_C0_USER_DATA_METADATA_MASK); MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0); MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, ft_attr.max_fte - 1); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c index 381325b4a863..00ff809dcfe8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c @@ -832,8 +832,7 @@ mlx5_chains_init(struct mlx5_core_dev *dev, struct mlx5_chains_attr *attr) if (err) goto init_prios_ht_err; - mapping = mapping_create(sizeof(u32), attr->max_restore_tag, - true); + mapping = mapping_create(sizeof(struct mlx5_mapped_obj), attr->max_restore_tag, true); if (IS_ERR(mapping)) { err = PTR_ERR(mapping); goto mapping_err; @@ -884,21 +883,28 @@ int mlx5_chains_get_chain_mapping(struct mlx5_fs_chains *chains, u32 chain, u32 *chain_mapping) { - return mapping_add(chains_mapping(chains), &chain, chain_mapping); + struct mapping_ctx *ctx = chains->chains_mapping; + struct mlx5_mapped_obj mapped_obj = {}; + + mapped_obj.type = MLX5_MAPPED_OBJ_CHAIN; + mapped_obj.chain = chain; + return mapping_add(ctx, &mapped_obj, chain_mapping); } int mlx5_chains_put_chain_mapping(struct mlx5_fs_chains *chains, u32 chain_mapping) { - return mapping_remove(chains_mapping(chains), chain_mapping); + struct mapping_ctx *ctx = chains->chains_mapping; + + return mapping_remove(ctx, chain_mapping); } -int mlx5_get_chain_for_tag(struct mlx5_fs_chains *chains, u32 tag, - u32 *chain) +int +mlx5_get_mapped_object(struct mlx5_fs_chains *chains, u32 tag, struct mlx5_mapped_obj *obj) { int err; - err = mapping_find(chains_mapping(chains), tag, chain); + err = mapping_find(chains->chains_mapping, tag, obj); if (err) { mlx5_core_warn(chains->dev, "Can't find chain for tag: %d\n", tag); return -ENOENT; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h index 6d5be31b05dd..75a3bba12a78 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h @@ -7,6 +7,7 @@ #include struct mlx5_fs_chains; +struct mlx5_mapped_obj; enum mlx5_chains_flags { MLX5_CHAINS_AND_PRIOS_SUPPORTED = BIT(0), @@ -64,7 +65,7 @@ mlx5_chains_create(struct mlx5_core_dev *dev, struct mlx5_chains_attr *attr); void mlx5_chains_destroy(struct mlx5_fs_chains *chains); int -mlx5_get_chain_for_tag(struct mlx5_fs_chains *chains, u32 tag, u32 *chain); +mlx5_get_mapped_object(struct mlx5_fs_chains *chains, u32 tag, struct mlx5_mapped_obj *obj); void mlx5_chains_set_end_ft(struct mlx5_fs_chains *chains, diff --git a/include/linux/mlx5/eswitch.h b/include/linux/mlx5/eswitch.h index 994c2c8cb4fd..125ae482383b 100644 --- a/include/linux/mlx5/eswitch.h +++ b/include/linux/mlx5/eswitch.h @@ -74,20 +74,19 @@ bool mlx5_eswitch_reg_c1_loopback_enabled(const struct mlx5_eswitch *esw); bool mlx5_eswitch_vport_match_metadata_enabled(const struct mlx5_eswitch *esw); /* Reg C0 usage: - * Reg C0 = < ESW_PFNUM_BITS(4) | ESW_VPORT BITS(12) | ESW_CHAIN_TAG(16) > + * Reg C0 = < ESW_PFNUM_BITS(4) | ESW_VPORT BITS(12) | ESW_REG_C0_OBJ(16) > * * Highest 4 bits of the reg c0 is the PF_NUM (range 0-15), 12 bits of * unique non-zero vport id (range 1-4095). The rest (lowest 16 bits) is left - * for tc chain tag restoration. + * for user data objects managed by a common mapping context. * PFNUM + VPORT comprise the SOURCE_PORT matching. */ #define ESW_VPORT_BITS 12 #define ESW_PFNUM_BITS 4 #define ESW_SOURCE_PORT_METADATA_BITS (ESW_PFNUM_BITS + ESW_VPORT_BITS) #define ESW_SOURCE_PORT_METADATA_OFFSET (32 - ESW_SOURCE_PORT_METADATA_BITS) -#define ESW_CHAIN_TAG_METADATA_BITS (32 - ESW_SOURCE_PORT_METADATA_BITS) -#define ESW_CHAIN_TAG_METADATA_MASK GENMASK(ESW_CHAIN_TAG_METADATA_BITS - 1,\ - 0) +#define ESW_REG_C0_USER_DATA_METADATA_BITS (32 - ESW_SOURCE_PORT_METADATA_BITS) +#define ESW_REG_C0_USER_DATA_METADATA_MASK GENMASK(ESW_REG_C0_USER_DATA_METADATA_BITS - 1, 0) static inline u32 mlx5_eswitch_get_vport_metadata_mask(void) { From c935568271b5b25baea65ad623f146de72570cd7 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Mon, 31 Aug 2020 13:27:53 +0800 Subject: [PATCH 06/13] net/mlx5: Instantiate separate mapping objects for FDB and NIC tables Currently, the u32 chain id is mapped to u16 value which is stored on the lower 16 bits of reg_c0 for FDB and reg_b for NIC tables. The mapping is internally maintained by the chains object. However, with the introduction of reg_c0 objects the fdb may store more than just the chain id on reg_c0. This is not relevant for NIC tables. Separate the chains mapping instantiation for FDB and NIC tables. Remove the mapping from the chains object. For FDB tables, create the mapping per eswitch. For NIC tables, create the mapping per tc table. Pass the corresponding mapping pointer when creating the chains object. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/Makefile | 5 ++- .../net/ethernet/mellanox/mlx5/core/en/fs.h | 1 + .../ethernet/mellanox/mlx5/core/en/rep/tc.c | 2 +- .../net/ethernet/mellanox/mlx5/core/en_tc.c | 19 +++++++-- .../net/ethernet/mellanox/mlx5/core/eswitch.h | 1 + .../mellanox/mlx5/core/eswitch_offloads.c | 16 ++++++- .../mellanox/mlx5/core/lib/fs_chains.c | 42 +++---------------- .../mellanox/mlx5/core/lib/fs_chains.h | 5 +-- 8 files changed, 43 insertions(+), 48 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index 4c86b68ad26c..5bedc2c4d26f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -37,9 +37,10 @@ mlx5_core-$(CONFIG_MLX5_EN_RXNFC) += en_fs_ethtool.o mlx5_core-$(CONFIG_MLX5_CORE_EN_DCB) += en_dcbnl.o en/port_buffer.o mlx5_core-$(CONFIG_PCI_HYPERV_INTERFACE) += en/hv_vhca_stats.o mlx5_core-$(CONFIG_MLX5_ESWITCH) += lag_mp.o lib/geneve.o lib/port_tun.o \ - en_rep.o en/rep/bond.o en/mod_hdr.o + en_rep.o en/rep/bond.o en/mod_hdr.o \ + en/mapping.o mlx5_core-$(CONFIG_MLX5_CLS_ACT) += en_tc.o en/rep/tc.o en/rep/neigh.o \ - en/mapping.o lib/fs_chains.o en/tc_tun.o \ + lib/fs_chains.o en/tc_tun.o \ esw/indir_table.o en/tc_tun_encap.o \ en/tc_tun_vxlan.o en/tc_tun_gre.o en/tc_tun_geneve.o \ en/tc_tun_mplsoudp.o diag/en_tc_tracepoint.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h index 373bd89484cb..1d5ce07b83f4 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h @@ -29,6 +29,7 @@ struct mlx5e_tc_table { struct netdev_net_notifier netdevice_nn; struct mlx5_tc_ct_priv *ct; + struct mapping_ctx *mapping; }; struct mlx5e_flow_table { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c index dde83cba85c3..9b55a5c394d0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c @@ -641,7 +641,7 @@ bool mlx5e_rep_tc_update_skb(struct mlx5_cqe64 *cqe, priv = netdev_priv(skb->dev); esw = priv->mdev->priv.eswitch; - err = mlx5_get_mapped_object(esw_chains(esw), reg_c0, &mapped_obj); + err = mapping_find(esw->offloads.reg_c0_obj_pool, reg_c0, &mapped_obj); if (err) { netdev_dbg(priv->netdev, "Couldn't find mapped object for reg_c0: %d, err: %d\n", diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 9b5607ddb9a2..c938215c8fbc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -4731,6 +4731,7 @@ int mlx5e_tc_nic_init(struct mlx5e_priv *priv) { struct mlx5e_tc_table *tc = &priv->fs.tc; struct mlx5_core_dev *dev = priv->mdev; + struct mapping_ctx *chains_mapping; struct mlx5_chains_attr attr = {}; int err; @@ -4745,15 +4746,22 @@ int mlx5e_tc_nic_init(struct mlx5e_priv *priv) lockdep_set_class(&tc->ht.mutex, &tc_ht_lock_key); - if (MLX5_CAP_FLOWTABLE_NIC_RX(priv->mdev, ignore_flow_level)) { + chains_mapping = mapping_create(sizeof(struct mlx5_mapped_obj), + MLX5E_TC_TABLE_CHAIN_TAG_MASK, true); + if (IS_ERR(chains_mapping)) { + err = PTR_ERR(chains_mapping); + goto err_mapping; + } + tc->mapping = chains_mapping; + + if (MLX5_CAP_FLOWTABLE_NIC_RX(priv->mdev, ignore_flow_level)) attr.flags = MLX5_CHAINS_AND_PRIOS_SUPPORTED | MLX5_CHAINS_IGNORE_FLOW_LEVEL_SUPPORTED; - attr.max_restore_tag = MLX5E_TC_TABLE_CHAIN_TAG_MASK; - } attr.ns = MLX5_FLOW_NAMESPACE_KERNEL; attr.max_ft_sz = mlx5e_tc_nic_get_ft_size(dev); attr.max_grp_num = MLX5E_TC_TABLE_NUM_GROUPS; attr.default_ft = mlx5e_vlan_get_flowtable(priv->fs.vlan); + attr.mapping = chains_mapping; tc->chains = mlx5_chains_create(dev, &attr); if (IS_ERR(tc->chains)) { @@ -4780,6 +4788,8 @@ err_reg: mlx5_tc_ct_clean(tc->ct); mlx5_chains_destroy(tc->chains); err_chains: + mapping_destroy(chains_mapping); +err_mapping: rhashtable_destroy(&tc->ht); return err; } @@ -4814,6 +4824,7 @@ void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv) mutex_destroy(&tc->t_lock); mlx5_tc_ct_clean(tc->ct); + mapping_destroy(tc->mapping); mlx5_chains_destroy(tc->chains); } @@ -4981,7 +4992,7 @@ bool mlx5e_tc_update_skb(struct mlx5_cqe64 *cqe, chain_tag = reg_b & MLX5E_TC_TABLE_CHAIN_TAG_MASK; - err = mlx5_get_mapped_object(nic_chains(priv), chain_tag, &mapped_obj); + err = mapping_find(tc->mapping, chain_tag, &mapped_obj); if (err) { netdev_dbg(priv->netdev, "Couldn't find chain for chain tag: %d, err: %d\n", diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index c5b35e7f8aed..a97396330160 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -217,6 +217,7 @@ struct mlx5_esw_offload { struct mlx5_flow_table *ft_offloads_restore; struct mlx5_flow_group *restore_group; struct mlx5_modify_hdr *restore_copy_hdr_id; + struct mapping_ctx *reg_c0_obj_pool; struct mlx5_flow_table *ft_offloads; struct mlx5_flow_group *vport_rx_group; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 117d9fa93ff5..510e9b8b24fe 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -47,6 +47,7 @@ #include "lib/eq.h" #include "lib/fs_chains.h" #include "en_tc.h" +#include "en/mapping.h" /* There are two match-all miss flows, one for unicast dst mac and * one for multicast. @@ -1428,7 +1429,7 @@ esw_chains_create(struct mlx5_eswitch *esw, struct mlx5_flow_table *miss_fdb) attr.max_ft_sz = fdb_max; attr.max_grp_num = esw->params.large_group_num; attr.default_ft = miss_fdb; - attr.max_restore_tag = ESW_REG_C0_USER_DATA_METADATA_MASK; + attr.mapping = esw->offloads.reg_c0_obj_pool; chains = mlx5_chains_create(dev, &attr); if (IS_ERR(chains)) { @@ -2595,6 +2596,7 @@ static int mlx5_esw_host_number_init(struct mlx5_eswitch *esw) int esw_offloads_enable(struct mlx5_eswitch *esw) { + struct mapping_ctx *reg_c0_obj_pool; struct mlx5_vport *vport; int err, i; @@ -2622,6 +2624,15 @@ int esw_offloads_enable(struct mlx5_eswitch *esw) if (err) goto err_vport_metadata; + reg_c0_obj_pool = mapping_create(sizeof(struct mlx5_mapped_obj), + ESW_REG_C0_USER_DATA_METADATA_MASK, + true); + if (IS_ERR(reg_c0_obj_pool)) { + err = PTR_ERR(reg_c0_obj_pool); + goto err_pool; + } + esw->offloads.reg_c0_obj_pool = reg_c0_obj_pool; + err = esw_offloads_steering_init(esw); if (err) goto err_steering_init; @@ -2648,6 +2659,8 @@ err_vports: err_uplink: esw_offloads_steering_cleanup(esw); err_steering_init: + mapping_destroy(reg_c0_obj_pool); +err_pool: esw_set_passing_vport_metadata(esw, false); err_vport_metadata: esw_offloads_metadata_uninit(esw); @@ -2686,6 +2699,7 @@ void esw_offloads_disable(struct mlx5_eswitch *esw) esw_offloads_unload_rep(esw, MLX5_VPORT_UPLINK); esw_set_passing_vport_metadata(esw, false); esw_offloads_steering_cleanup(esw); + mapping_destroy(esw->offloads.reg_c0_obj_pool); esw_offloads_metadata_uninit(esw); esw->flags &= ~MLX5_ESWITCH_VPORT_MATCH_METADATA; mlx5_rdma_disable_roce(esw->dev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c index 00ff809dcfe8..00ef10a1a9f8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c @@ -7,15 +7,11 @@ #include "lib/fs_chains.h" #include "en/mapping.h" -#include "mlx5_core.h" #include "fs_core.h" -#include "eswitch.h" -#include "en.h" #include "en_tc.h" #define chains_lock(chains) ((chains)->lock) #define chains_ht(chains) ((chains)->chains_ht) -#define chains_mapping(chains) ((chains)->chains_mapping) #define prios_ht(chains) ((chains)->prios_ht) #define ft_pool_left(chains) ((chains)->ft_left) #define tc_default_ft(chains) ((chains)->tc_default_ft) @@ -300,7 +296,7 @@ create_chain_restore(struct fs_chain *chain) !mlx5_chains_prios_supported(chains)) return 0; - err = mapping_add(chains_mapping(chains), &chain->chain, &index); + err = mlx5_chains_get_chain_mapping(chains, chain->chain, &index); if (err) return err; if (index == MLX5_FS_DEFAULT_FLOW_TAG) { @@ -310,10 +306,8 @@ create_chain_restore(struct fs_chain *chain) * * This case isn't possible with MLX5_FS_DEFAULT_FLOW_TAG = 0. */ - err = mapping_add(chains_mapping(chains), - &chain->chain, &index); - mapping_remove(chains_mapping(chains), - MLX5_FS_DEFAULT_FLOW_TAG); + err = mlx5_chains_get_chain_mapping(chains, chain->chain, &index); + mapping_remove(chains->chains_mapping, MLX5_FS_DEFAULT_FLOW_TAG); if (err) return err; } @@ -361,7 +355,7 @@ err_mod_hdr: mlx5_del_flow_rules(chain->restore_rule); err_rule: /* Datapath can't find this mapping, so we can safely remove it */ - mapping_remove(chains_mapping(chains), chain->id); + mapping_remove(chains->chains_mapping, chain->id); return err; } @@ -376,7 +370,7 @@ static void destroy_chain_restore(struct fs_chain *chain) mlx5_del_flow_rules(chain->restore_rule); mlx5_modify_header_dealloc(chains->dev, chain->miss_modify_hdr); - mapping_remove(chains_mapping(chains), chain->id); + mapping_remove(chains->chains_mapping, chain->id); } static struct fs_chain * @@ -797,7 +791,6 @@ static struct mlx5_fs_chains * mlx5_chains_init(struct mlx5_core_dev *dev, struct mlx5_chains_attr *attr) { struct mlx5_fs_chains *chains_priv; - struct mapping_ctx *mapping; u32 max_flow_counter; int err; @@ -816,6 +809,7 @@ mlx5_chains_init(struct mlx5_core_dev *dev, struct mlx5_chains_attr *attr) chains_priv->flags = attr->flags; chains_priv->ns = attr->ns; chains_priv->group_num = attr->max_grp_num; + chains_priv->chains_mapping = attr->mapping; tc_default_ft(chains_priv) = tc_end_ft(chains_priv) = attr->default_ft; mlx5_core_info(dev, "Supported tc offload range - chains: %u, prios: %u\n", @@ -832,19 +826,10 @@ mlx5_chains_init(struct mlx5_core_dev *dev, struct mlx5_chains_attr *attr) if (err) goto init_prios_ht_err; - mapping = mapping_create(sizeof(struct mlx5_mapped_obj), attr->max_restore_tag, true); - if (IS_ERR(mapping)) { - err = PTR_ERR(mapping); - goto mapping_err; - } - chains_mapping(chains_priv) = mapping; - mutex_init(&chains_lock(chains_priv)); return chains_priv; -mapping_err: - rhashtable_destroy(&prios_ht(chains_priv)); init_prios_ht_err: rhashtable_destroy(&chains_ht(chains_priv)); init_chains_ht_err: @@ -856,7 +841,6 @@ static void mlx5_chains_cleanup(struct mlx5_fs_chains *chains) { mutex_destroy(&chains_lock(chains)); - mapping_destroy(chains_mapping(chains)); rhashtable_destroy(&prios_ht(chains)); rhashtable_destroy(&chains_ht(chains)); @@ -898,17 +882,3 @@ mlx5_chains_put_chain_mapping(struct mlx5_fs_chains *chains, u32 chain_mapping) return mapping_remove(ctx, chain_mapping); } - -int -mlx5_get_mapped_object(struct mlx5_fs_chains *chains, u32 tag, struct mlx5_mapped_obj *obj) -{ - int err; - - err = mapping_find(chains->chains_mapping, tag, obj); - if (err) { - mlx5_core_warn(chains->dev, "Can't find chain for tag: %d\n", tag); - return -ENOENT; - } - - return 0; -} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h index 75a3bba12a78..e96f345e7dae 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.h @@ -21,7 +21,7 @@ struct mlx5_chains_attr { u32 max_ft_sz; u32 max_grp_num; struct mlx5_flow_table *default_ft; - u32 max_restore_tag; + struct mapping_ctx *mapping; }; #if IS_ENABLED(CONFIG_MLX5_CLS_ACT) @@ -64,9 +64,6 @@ struct mlx5_fs_chains * mlx5_chains_create(struct mlx5_core_dev *dev, struct mlx5_chains_attr *attr); void mlx5_chains_destroy(struct mlx5_fs_chains *chains); -int -mlx5_get_mapped_object(struct mlx5_fs_chains *chains, u32 tag, struct mlx5_mapped_obj *obj); - void mlx5_chains_set_end_ft(struct mlx5_fs_chains *chains, struct mlx5_flow_table *ft); From 41c2fd949803a5ff8acfed1d81c0bbd62d8f660d Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Mon, 31 Aug 2020 13:28:35 +0800 Subject: [PATCH 07/13] net/mlx5e: TC, Parse sample action Parse TC sample action and save sample parameters in flow attribute data structure. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/en/tc_priv.h | 1 + .../net/ethernet/mellanox/mlx5/core/en_tc.c | 28 +++++++++++++++++++ .../ethernet/mellanox/mlx5/core/esw/sample.h | 13 +++++++++ .../net/ethernet/mellanox/mlx5/core/eswitch.h | 2 ++ 4 files changed, 44 insertions(+) create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h index c223591ffc22..d1599b7b944b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h @@ -27,6 +27,7 @@ enum { MLX5E_TC_FLOW_FLAG_L3_TO_L2_DECAP = MLX5E_TC_FLOW_BASE + 8, MLX5E_TC_FLOW_FLAG_TUN_RX = MLX5E_TC_FLOW_BASE + 9, MLX5E_TC_FLOW_FLAG_FAILED = MLX5E_TC_FLOW_BASE + 10, + MLX5E_TC_FLOW_FLAG_SAMPLE = MLX5E_TC_FLOW_BASE + 11, }; struct mlx5e_tc_flow_parse_attr { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index c938215c8fbc..85782d12ffb2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -47,6 +47,7 @@ #include #include #include +#include #include #include #include @@ -1481,6 +1482,7 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv, if (flow_flag_test(flow, L3_TO_L2_DECAP)) mlx5e_detach_decap(priv, flow); + kfree(flow->attr->esw_attr->sample); kfree(flow->attr); } @@ -3627,6 +3629,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, bool ft_flow = mlx5e_is_ft_flow(flow); const struct flow_action_entry *act; struct mlx5_esw_flow_attr *esw_attr; + struct mlx5_sample_attr sample = {}; bool encap = false, decap = false; u32 action = attr->action; int err, i, if_count = 0; @@ -3881,6 +3884,10 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, attr->dest_chain = act->chain_index; break; case FLOW_ACTION_CT: + if (flow_flag_test(flow, SAMPLE)) { + NL_SET_ERR_MSG_MOD(extack, "Sample action with connection tracking is not supported"); + return -EOPNOTSUPP; + } err = mlx5_tc_ct_parse_action(get_ct_priv(priv), attr, act, extack); if (err) return err; @@ -3888,6 +3895,17 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, flow_flag_set(flow, CT); esw_attr->split_count = esw_attr->out_count; break; + case FLOW_ACTION_SAMPLE: + if (flow_flag_test(flow, CT)) { + NL_SET_ERR_MSG_MOD(extack, "Sample action with connection tracking is not supported"); + return -EOPNOTSUPP; + } + sample.rate = act->sample.rate; + sample.group_num = act->sample.psample_group->group_num; + if (act->sample.truncate) + sample.trunc_size = act->sample.trunc_size; + flow_flag_set(flow, SAMPLE); + break; default: NL_SET_ERR_MSG_MOD(extack, "The offload action is not supported"); return -EOPNOTSUPP; @@ -3966,6 +3984,16 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, return -EOPNOTSUPP; } + /* Allocate sample attribute only when there is a sample action and + * no errors after parsing. + */ + if (flow_flag_test(flow, SAMPLE)) { + esw_attr->sample = kzalloc(sizeof(*esw_attr->sample), GFP_KERNEL); + if (!esw_attr->sample) + return -ENOMEM; + *esw_attr->sample = sample; + } + return 0; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h new file mode 100644 index 000000000000..35a5e6dddcd0 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h @@ -0,0 +1,13 @@ +/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */ +/* Copyright (c) 2021 Mellanox Technologies. */ + +#ifndef __MLX5_EN_TC_SAMPLE_H__ +#define __MLX5_EN_TC_SAMPLE_H__ + +struct mlx5_sample_attr { + u32 group_num; + u32 rate; + u32 trunc_size; +}; + +#endif /* __MLX5_EN_TC_SAMPLE_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index a97396330160..9b26bd67e2b8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -46,6 +46,7 @@ #include "lib/fs_chains.h" #include "sf/sf.h" #include "en/tc_ct.h" +#include "esw/sample.h" enum mlx5_mapped_obj_type { MLX5_MAPPED_OBJ_CHAIN, @@ -440,6 +441,7 @@ struct mlx5_esw_flow_attr { } dests[MLX5_MAX_FLOW_FWD_VPORTS]; struct mlx5_rx_tun_attr *rx_tun_attr; struct mlx5_pkt_reformat *decap_pkt_reformat; + struct mlx5_sample_attr *sample; }; int mlx5_devlink_eswitch_mode_set(struct devlink *devlink, u16 mode, From 2a9ab10a5689a4612d441df1cc628c381dc75ed3 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Fri, 18 Sep 2020 17:46:57 +0800 Subject: [PATCH 08/13] net/mlx5e: TC, Add sampler termination table API Sampled packets are sent to software using termination tables. There is only one rule in that table that is to forward sampled packets to the e-switch management vport. Create a sampler termination table and rule for each eswitch. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/Kconfig | 12 +++ .../net/ethernet/mellanox/mlx5/core/Makefile | 1 + .../net/ethernet/mellanox/mlx5/core/en_rep.h | 1 + .../net/ethernet/mellanox/mlx5/core/en_tc.c | 11 +++ .../ethernet/mellanox/mlx5/core/esw/sample.c | 96 +++++++++++++++++++ .../ethernet/mellanox/mlx5/core/esw/sample.h | 8 ++ 6 files changed, 129 insertions(+) create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig index 9d623e38d783..461a43f338e6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig +++ b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig @@ -104,6 +104,18 @@ config MLX5_TC_CT If unsure, set to Y +config MLX5_TC_SAMPLE + bool "MLX5 TC sample offload support" + depends on MLX5_CLS_ACT + default y + help + Say Y here if you want to support offloading sample rules via tc + sample action. + If set to N, will not be able to configure tc rules with sample + action. + + If unsure, set to Y + config MLX5_CORE_EN_DCB bool "Data Center Bridging (DCB) Support" default y diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index 5bedc2c4d26f..8bde58379ac6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -55,6 +55,7 @@ mlx5_core-$(CONFIG_MLX5_ESWITCH) += esw/acl/helper.o \ esw/acl/egress_lgcy.o esw/acl/egress_ofld.o \ esw/acl/ingress_lgcy.o esw/acl/ingress_ofld.o \ esw/devlink_port.o esw/vporttbl.o +mlx5_core-$(CONFIG_MLX5_TC_SAMPLE) += esw/sample.o mlx5_core-$(CONFIG_MLX5_MPFS) += lib/mpfs.o mlx5_core-$(CONFIG_VXLAN) += lib/vxlan.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h index 931fa619cb01..22585015c7a7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h @@ -89,6 +89,7 @@ struct mlx5_rep_uplink_priv { struct mapping_ctx *tunnel_enc_opts_mapping; struct mlx5_tc_ct_priv *ct_priv; + struct mlx5_esw_psample *esw_psample; /* support eswitch vports bonding */ struct mlx5e_rep_bond *bond; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 85782d12ffb2..1a403112defd 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -66,6 +66,7 @@ #include "en/mod_hdr.h" #include "en/tc_priv.h" #include "en/tc_tun_encap.h" +#include "esw/sample.h" #include "lib/devcom.h" #include "lib/geneve.h" #include "lib/fs_chains.h" @@ -4876,6 +4877,10 @@ int mlx5e_tc_esw_init(struct rhashtable *tc_ht) &esw->offloads.mod_hdr, MLX5_FLOW_NAMESPACE_FDB); +#if IS_ENABLED(CONFIG_MLX5_TC_SAMPLE) + uplink_priv->esw_psample = mlx5_esw_sample_init(netdev_priv(priv->netdev)); +#endif + mapping = mapping_create(sizeof(struct tunnel_match_key), TUNNEL_INFO_BITS_MASK, true); if (IS_ERR(mapping)) { @@ -4913,6 +4918,9 @@ err_ht_init: err_enc_opts_mapping: mapping_destroy(uplink_priv->tunnel_mapping); err_tun_mapping: +#if IS_ENABLED(CONFIG_MLX5_TC_SAMPLE) + mlx5_esw_sample_cleanup(uplink_priv->esw_psample); +#endif mlx5_tc_ct_clean(uplink_priv->ct_priv); netdev_warn(priv->netdev, "Failed to initialize tc (eswitch), err: %d", err); @@ -4931,6 +4939,9 @@ void mlx5e_tc_esw_cleanup(struct rhashtable *tc_ht) mapping_destroy(uplink_priv->tunnel_enc_opts_mapping); mapping_destroy(uplink_priv->tunnel_mapping); +#if IS_ENABLED(CONFIG_MLX5_TC_SAMPLE) + mlx5_esw_sample_cleanup(uplink_priv->esw_psample); +#endif mlx5_tc_ct_clean(uplink_priv->ct_priv); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c new file mode 100644 index 000000000000..9bd996e8d28a --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c @@ -0,0 +1,96 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +/* Copyright (c) 2021 Mellanox Technologies. */ + +#include "esw/sample.h" +#include "eswitch.h" + +struct mlx5_esw_psample { + struct mlx5e_priv *priv; + struct mlx5_flow_table *termtbl; + struct mlx5_flow_handle *termtbl_rule; +}; + +static int +sampler_termtbl_create(struct mlx5_esw_psample *esw_psample) +{ + struct mlx5_core_dev *dev = esw_psample->priv->mdev; + struct mlx5_eswitch *esw = dev->priv.eswitch; + struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5_flow_destination dest = {}; + struct mlx5_flow_namespace *root_ns; + struct mlx5_flow_act act = {}; + int err; + + if (!MLX5_CAP_ESW_FLOWTABLE_FDB(dev, termination_table)) { + mlx5_core_warn(dev, "termination table is not supported\n"); + return -EOPNOTSUPP; + } + + root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_FDB); + if (!root_ns) { + mlx5_core_warn(dev, "failed to get FDB flow namespace\n"); + return -EOPNOTSUPP; + } + + ft_attr.flags = MLX5_FLOW_TABLE_TERMINATION | MLX5_FLOW_TABLE_UNMANAGED; + ft_attr.autogroup.max_num_groups = 1; + ft_attr.prio = FDB_SLOW_PATH; + ft_attr.max_fte = 1; + ft_attr.level = 1; + esw_psample->termtbl = mlx5_create_auto_grouped_flow_table(root_ns, &ft_attr); + if (IS_ERR(esw_psample->termtbl)) { + err = PTR_ERR(esw_psample->termtbl); + mlx5_core_warn(dev, "failed to create termtbl, err: %d\n", err); + return err; + } + + act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; + dest.vport.num = esw->manager_vport; + esw_psample->termtbl_rule = mlx5_add_flow_rules(esw_psample->termtbl, NULL, &act, &dest, 1); + if (IS_ERR(esw_psample->termtbl_rule)) { + err = PTR_ERR(esw_psample->termtbl_rule); + mlx5_core_warn(dev, "failed to create termtbl rule, err: %d\n", err); + mlx5_destroy_flow_table(esw_psample->termtbl); + return err; + } + + return 0; +} + +static void +sampler_termtbl_destroy(struct mlx5_esw_psample *esw_psample) +{ + mlx5_del_flow_rules(esw_psample->termtbl_rule); + mlx5_destroy_flow_table(esw_psample->termtbl); +} + +struct mlx5_esw_psample * +mlx5_esw_sample_init(struct mlx5e_priv *priv) +{ + struct mlx5_esw_psample *esw_psample; + int err; + + esw_psample = kzalloc(sizeof(*esw_psample), GFP_KERNEL); + if (!esw_psample) + return ERR_PTR(-ENOMEM); + esw_psample->priv = priv; + err = sampler_termtbl_create(esw_psample); + if (err) + goto err_termtbl; + + return esw_psample; + +err_termtbl: + kfree(esw_psample); + return ERR_PTR(err); +} + +void +mlx5_esw_sample_cleanup(struct mlx5_esw_psample *esw_psample) +{ + if (IS_ERR_OR_NULL(esw_psample)) + return; + + sampler_termtbl_destroy(esw_psample); + kfree(esw_psample); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h index 35a5e6dddcd0..e42e3cb01c8c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h @@ -4,10 +4,18 @@ #ifndef __MLX5_EN_TC_SAMPLE_H__ #define __MLX5_EN_TC_SAMPLE_H__ +#include "en.h" + struct mlx5_sample_attr { u32 group_num; u32 rate; u32 trunc_size; }; +struct mlx5_esw_psample * +mlx5_esw_sample_init(struct mlx5e_priv *priv); + +void +mlx5_esw_sample_cleanup(struct mlx5_esw_psample *esw_psample); + #endif /* __MLX5_EN_TC_SAMPLE_H__ */ From 11ecd6c60b4ee6f5d5bcd91f5b87e50bad8f142a Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Mon, 21 Sep 2020 13:25:54 +0800 Subject: [PATCH 09/13] net/mlx5e: TC, Add sampler object API In order to offload sample action, HW introduces sampler object. The sampler object samples packets according to the provided sample ratio. Sampled packets are duplicated. One copy is processed by a termination table, named the sample table, which sends the packet up to software. The second copy is processed by the default table. Instantiate sampler object. Re-use identical sampler object for the same sample ratio, sample table and default table as a prestep for offloading tc sample actions. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/esw/sample.c | 131 ++++++++++++++++++ 1 file changed, 131 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c index 9bd996e8d28a..37e33670bb24 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c @@ -3,11 +3,28 @@ #include "esw/sample.h" #include "eswitch.h" +#include "en_tc.h" +#include "fs_core.h" struct mlx5_esw_psample { struct mlx5e_priv *priv; struct mlx5_flow_table *termtbl; struct mlx5_flow_handle *termtbl_rule; + DECLARE_HASHTABLE(hashtbl, 8); + struct mutex ht_lock; /* protect hashtbl */ +}; + +struct mlx5_sampler { + struct hlist_node hlist; + u32 sampler_id; + u32 sample_ratio; + u32 sample_table_id; + u32 default_table_id; + int count; +}; + +struct mlx5_sample_flow { + struct mlx5_sampler *sampler; }; static int @@ -64,6 +81,117 @@ sampler_termtbl_destroy(struct mlx5_esw_psample *esw_psample) mlx5_destroy_flow_table(esw_psample->termtbl); } +static int +sampler_obj_create(struct mlx5_core_dev *mdev, struct mlx5_sampler *sampler) +{ + u32 in[MLX5_ST_SZ_DW(create_sampler_obj_in)] = {}; + u32 out[MLX5_ST_SZ_DW(general_obj_out_cmd_hdr)]; + u64 general_obj_types; + void *obj; + int err; + + general_obj_types = MLX5_CAP_GEN_64(mdev, general_obj_types); + if (!(general_obj_types & MLX5_HCA_CAP_GENERAL_OBJECT_TYPES_SAMPLER)) + return -EOPNOTSUPP; + if (!MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, ignore_flow_level)) + return -EOPNOTSUPP; + + obj = MLX5_ADDR_OF(create_sampler_obj_in, in, sampler_object); + MLX5_SET(sampler_obj, obj, table_type, FS_FT_FDB); + MLX5_SET(sampler_obj, obj, ignore_flow_level, 1); + MLX5_SET(sampler_obj, obj, level, 1); + MLX5_SET(sampler_obj, obj, sample_ratio, sampler->sample_ratio); + MLX5_SET(sampler_obj, obj, sample_table_id, sampler->sample_table_id); + MLX5_SET(sampler_obj, obj, default_table_id, sampler->default_table_id); + MLX5_SET(general_obj_in_cmd_hdr, in, opcode, MLX5_CMD_OP_CREATE_GENERAL_OBJECT); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_type, MLX5_GENERAL_OBJECT_TYPES_SAMPLER); + + err = mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); + if (!err) + sampler->sampler_id = MLX5_GET(general_obj_out_cmd_hdr, out, obj_id); + + return err; +} + +static void +sampler_obj_destroy(struct mlx5_core_dev *mdev, u32 sampler_id) +{ + u32 in[MLX5_ST_SZ_DW(general_obj_in_cmd_hdr)] = {}; + u32 out[MLX5_ST_SZ_DW(general_obj_out_cmd_hdr)]; + + MLX5_SET(general_obj_in_cmd_hdr, in, opcode, MLX5_CMD_OP_DESTROY_GENERAL_OBJECT); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_type, MLX5_GENERAL_OBJECT_TYPES_SAMPLER); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_id, sampler_id); + + mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); +} + +static u32 +sampler_hash(u32 sample_ratio, u32 default_table_id) +{ + return jhash_2words(sample_ratio, default_table_id, 0); +} + +static int +sampler_cmp(u32 sample_ratio1, u32 default_table_id1, u32 sample_ratio2, u32 default_table_id2) +{ + return sample_ratio1 != sample_ratio2 || default_table_id1 != default_table_id2; +} + +static struct mlx5_sampler * +sampler_get(struct mlx5_esw_psample *esw_psample, u32 sample_ratio, u32 default_table_id) +{ + struct mlx5_sampler *sampler; + u32 hash_key; + int err; + + mutex_lock(&esw_psample->ht_lock); + hash_key = sampler_hash(sample_ratio, default_table_id); + hash_for_each_possible(esw_psample->hashtbl, sampler, hlist, hash_key) + if (!sampler_cmp(sampler->sample_ratio, sampler->default_table_id, + sample_ratio, default_table_id)) + goto add_ref; + + sampler = kzalloc(sizeof(*sampler), GFP_KERNEL); + if (!sampler) { + err = -ENOMEM; + goto err_alloc; + } + + sampler->sample_table_id = esw_psample->termtbl->id; + sampler->default_table_id = default_table_id; + sampler->sample_ratio = sample_ratio; + + err = sampler_obj_create(esw_psample->priv->mdev, sampler); + if (err) + goto err_create; + + hash_add(esw_psample->hashtbl, &sampler->hlist, hash_key); + +add_ref: + sampler->count++; + mutex_unlock(&esw_psample->ht_lock); + return sampler; + +err_create: + kfree(sampler); +err_alloc: + mutex_unlock(&esw_psample->ht_lock); + return ERR_PTR(err); +} + +static void +sampler_put(struct mlx5_esw_psample *esw_psample, struct mlx5_sampler *sampler) +{ + mutex_lock(&esw_psample->ht_lock); + if (--sampler->count == 0) { + hash_del(&sampler->hlist); + sampler_obj_destroy(esw_psample->priv->mdev, sampler->sampler_id); + kfree(sampler); + } + mutex_unlock(&esw_psample->ht_lock); +} + struct mlx5_esw_psample * mlx5_esw_sample_init(struct mlx5e_priv *priv) { @@ -78,6 +206,8 @@ mlx5_esw_sample_init(struct mlx5e_priv *priv) if (err) goto err_termtbl; + mutex_init(&esw_psample->ht_lock); + return esw_psample; err_termtbl: @@ -91,6 +221,7 @@ mlx5_esw_sample_cleanup(struct mlx5_esw_psample *esw_psample) if (IS_ERR_OR_NULL(esw_psample)) return; + mutex_destroy(&esw_psample->ht_lock); sampler_termtbl_destroy(esw_psample); kfree(esw_psample); } From 36a3196256bf3310e5e7142b0e61787f7a201abd Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Tue, 26 Jan 2021 11:08:28 +0800 Subject: [PATCH 10/13] net/mlx5e: TC, Add sampler restore handle API Use common object pool to create an object ID to map sample parameters. Allocate a modify header action to write the object ID to reg_c0 lower 16 bits. Create a restore rule to pass the object ID to software. So software can identify sampled packets via the object ID and send it to userspace. Aggregate the modify header action, restore rule and object ID to a sample restore handle. Re-use identical sample restore handle for the same object ID. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/esw/sample.c | 108 ++++++++++++++++++ 1 file changed, 108 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c index 37e33670bb24..79a0e49b2799 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c @@ -1,6 +1,8 @@ // SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB /* Copyright (c) 2021 Mellanox Technologies. */ +#include +#include #include "esw/sample.h" #include "eswitch.h" #include "en_tc.h" @@ -12,6 +14,8 @@ struct mlx5_esw_psample { struct mlx5_flow_handle *termtbl_rule; DECLARE_HASHTABLE(hashtbl, 8); struct mutex ht_lock; /* protect hashtbl */ + DECLARE_HASHTABLE(restore_hashtbl, 8); + struct mutex restore_lock; /* protect restore_hashtbl */ }; struct mlx5_sampler { @@ -25,6 +29,15 @@ struct mlx5_sampler { struct mlx5_sample_flow { struct mlx5_sampler *sampler; + struct mlx5_sample_restore *restore; +}; + +struct mlx5_sample_restore { + struct hlist_node hlist; + struct mlx5_modify_hdr *modify_hdr; + struct mlx5_flow_handle *rule; + u32 obj_id; + int count; }; static int @@ -192,6 +205,99 @@ sampler_put(struct mlx5_esw_psample *esw_psample, struct mlx5_sampler *sampler) mutex_unlock(&esw_psample->ht_lock); } +static struct mlx5_modify_hdr * +sample_metadata_rule_get(struct mlx5_core_dev *mdev, u32 obj_id) +{ + struct mlx5e_tc_mod_hdr_acts mod_acts = {}; + struct mlx5_modify_hdr *modify_hdr; + int err; + + err = mlx5e_tc_match_to_reg_set(mdev, &mod_acts, MLX5_FLOW_NAMESPACE_FDB, + CHAIN_TO_REG, obj_id); + if (err) + goto err_set_regc0; + + modify_hdr = mlx5_modify_header_alloc(mdev, MLX5_FLOW_NAMESPACE_FDB, + mod_acts.num_actions, + mod_acts.actions); + if (IS_ERR(modify_hdr)) { + err = PTR_ERR(modify_hdr); + goto err_modify_hdr; + } + + dealloc_mod_hdr_actions(&mod_acts); + return modify_hdr; + +err_modify_hdr: + dealloc_mod_hdr_actions(&mod_acts); +err_set_regc0: + return ERR_PTR(err); +} + +static struct mlx5_sample_restore * +sample_restore_get(struct mlx5_esw_psample *esw_psample, u32 obj_id) +{ + struct mlx5_core_dev *mdev = esw_psample->priv->mdev; + struct mlx5_eswitch *esw = mdev->priv.eswitch; + struct mlx5_sample_restore *restore; + struct mlx5_modify_hdr *modify_hdr; + int err; + + mutex_lock(&esw_psample->restore_lock); + hash_for_each_possible(esw_psample->restore_hashtbl, restore, hlist, obj_id) + if (restore->obj_id == obj_id) + goto add_ref; + + restore = kzalloc(sizeof(*restore), GFP_KERNEL); + if (!restore) { + err = -ENOMEM; + goto err_alloc; + } + restore->obj_id = obj_id; + + modify_hdr = sample_metadata_rule_get(mdev, obj_id); + if (IS_ERR(modify_hdr)) { + err = PTR_ERR(modify_hdr); + goto err_modify_hdr; + } + restore->modify_hdr = modify_hdr; + + restore->rule = esw_add_restore_rule(esw, obj_id); + if (IS_ERR(restore->rule)) { + err = PTR_ERR(restore->rule); + goto err_restore; + } + + hash_add(esw_psample->restore_hashtbl, &restore->hlist, obj_id); +add_ref: + restore->count++; + mutex_unlock(&esw_psample->restore_lock); + return restore; + +err_restore: + mlx5_modify_header_dealloc(mdev, restore->modify_hdr); +err_modify_hdr: + kfree(restore); +err_alloc: + mutex_unlock(&esw_psample->restore_lock); + return ERR_PTR(err); +} + +static void +sample_restore_put(struct mlx5_esw_psample *esw_psample, struct mlx5_sample_restore *restore) +{ + mutex_lock(&esw_psample->restore_lock); + if (--restore->count == 0) + hash_del(&restore->hlist); + mutex_unlock(&esw_psample->restore_lock); + + if (!restore->count) { + mlx5_del_flow_rules(restore->rule); + mlx5_modify_header_dealloc(esw_psample->priv->mdev, restore->modify_hdr); + kfree(restore); + } +} + struct mlx5_esw_psample * mlx5_esw_sample_init(struct mlx5e_priv *priv) { @@ -207,6 +313,7 @@ mlx5_esw_sample_init(struct mlx5e_priv *priv) goto err_termtbl; mutex_init(&esw_psample->ht_lock); + mutex_init(&esw_psample->restore_lock); return esw_psample; @@ -221,6 +328,7 @@ mlx5_esw_sample_cleanup(struct mlx5_esw_psample *esw_psample) if (IS_ERR_OR_NULL(esw_psample)) return; + mutex_destroy(&esw_psample->restore_lock); mutex_destroy(&esw_psample->ht_lock); sampler_termtbl_destroy(esw_psample); kfree(esw_psample); From 7319a1cc3ca9f719e4ef1b1882f6fdb57d0719b8 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Fri, 15 Jan 2021 16:03:47 +0800 Subject: [PATCH 11/13] net/mlx5e: TC, Refactor tc update skb function As a pre-step to process sampled packet in this function. Signed-off-by: Chris Mi Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/en/rep/tc.c | 59 +++++++++++-------- 1 file changed, 33 insertions(+), 26 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c index 9b55a5c394d0..a25ec309d298 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c @@ -611,20 +611,46 @@ static bool mlx5e_restore_tunnel(struct mlx5e_priv *priv, struct sk_buff *skb, return true; } + +static bool mlx5e_restore_skb(struct sk_buff *skb, u32 chain, u32 reg_c1, + struct mlx5e_tc_update_priv *tc_priv) +{ + struct mlx5e_priv *priv = netdev_priv(skb->dev); + u32 tunnel_id = reg_c1 >> ESW_TUN_OFFSET; + + if (chain) { + struct mlx5_rep_uplink_priv *uplink_priv; + struct mlx5e_rep_priv *uplink_rpriv; + struct tc_skb_ext *tc_skb_ext; + struct mlx5_eswitch *esw; + u32 zone_restore_id; + + tc_skb_ext = skb_ext_add(skb, TC_SKB_EXT); + if (!tc_skb_ext) { + WARN_ON(1); + return false; + } + tc_skb_ext->chain = chain; + zone_restore_id = reg_c1 & ESW_ZONE_ID_MASK; + esw = priv->mdev->priv.eswitch; + uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); + uplink_priv = &uplink_rpriv->uplink_priv; + if (!mlx5e_tc_ct_restore_flow(uplink_priv->ct_priv, skb, + zone_restore_id)) + return false; + } + return mlx5e_restore_tunnel(priv, skb, tc_priv, tunnel_id); +} #endif /* CONFIG_NET_TC_SKB_EXT */ bool mlx5e_rep_tc_update_skb(struct mlx5_cqe64 *cqe, struct sk_buff *skb, struct mlx5e_tc_update_priv *tc_priv) { -#if IS_ENABLED(CONFIG_NET_TC_SKB_EXT) - u32 reg_c0, reg_c1, tunnel_id, zone_restore_id; - struct mlx5_rep_uplink_priv *uplink_priv; - struct mlx5e_rep_priv *uplink_rpriv; struct mlx5_mapped_obj mapped_obj; - struct tc_skb_ext *tc_skb_ext; struct mlx5_eswitch *esw; struct mlx5e_priv *priv; + u32 reg_c0, reg_c1; int err; reg_c0 = (be32_to_cpu(cqe->sop_drop_qpn) & MLX5E_TC_FLOW_ID_MASK); @@ -640,7 +666,6 @@ bool mlx5e_rep_tc_update_skb(struct mlx5_cqe64 *cqe, priv = netdev_priv(skb->dev); esw = priv->mdev->priv.eswitch; - err = mapping_find(esw->offloads.reg_c0_obj_pool, reg_c0, &mapped_obj); if (err) { netdev_dbg(priv->netdev, @@ -649,31 +674,13 @@ bool mlx5e_rep_tc_update_skb(struct mlx5_cqe64 *cqe, return false; } +#if IS_ENABLED(CONFIG_NET_TC_SKB_EXT) if (mapped_obj.type == MLX5_MAPPED_OBJ_CHAIN) { - if (mapped_obj.chain) { - tc_skb_ext = skb_ext_add(skb, TC_SKB_EXT); - if (!tc_skb_ext) { - WARN_ON(1); - return false; - } - - tc_skb_ext->chain = mapped_obj.chain; - - zone_restore_id = reg_c1 & ESW_ZONE_ID_MASK; - - uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); - uplink_priv = &uplink_rpriv->uplink_priv; - if (!mlx5e_tc_ct_restore_flow(uplink_priv->ct_priv, skb, - zone_restore_id)) - return false; - } + return mlx5e_restore_skb(skb, mapped_obj.chain, reg_c1, tc_priv); } else { netdev_dbg(priv->netdev, "Invalid mapped object type: %d\n", mapped_obj.type); return false; } - - tunnel_id = reg_c1 >> ESW_TUN_OFFSET; - return mlx5e_restore_tunnel(priv, skb, tc_priv, tunnel_id); #endif /* CONFIG_NET_TC_SKB_EXT */ return true; From be9dc0047450f7d8c0462a6ef3f7c26bced16911 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Tue, 26 Jan 2021 11:15:46 +0800 Subject: [PATCH 12/13] net/mlx5e: TC, Handle sampled packets Mark the sampled packets with a sample restore object. Send sampled packets using the psample api. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en/rep/tc.c | 14 +++++++++++--- .../net/ethernet/mellanox/mlx5/core/esw/sample.c | 15 +++++++++++++++ .../net/ethernet/mellanox/mlx5/core/esw/sample.h | 3 +++ drivers/net/ethernet/mellanox/mlx5/core/eswitch.h | 6 ++++++ 4 files changed, 35 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c index a25ec309d298..6cdc52d50a48 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c @@ -17,6 +17,7 @@ #include "en/mapping.h" #include "en/tc_tun.h" #include "lib/port_tun.h" +#include "esw/sample.h" struct mlx5e_rep_indr_block_priv { struct net_device *netdev; @@ -675,13 +676,20 @@ bool mlx5e_rep_tc_update_skb(struct mlx5_cqe64 *cqe, } #if IS_ENABLED(CONFIG_NET_TC_SKB_EXT) - if (mapped_obj.type == MLX5_MAPPED_OBJ_CHAIN) { + if (mapped_obj.type == MLX5_MAPPED_OBJ_CHAIN) return mlx5e_restore_skb(skb, mapped_obj.chain, reg_c1, tc_priv); - } else { +#endif /* CONFIG_NET_TC_SKB_EXT */ +#if IS_ENABLED(CONFIG_MLX5_TC_SAMPLE) + if (mapped_obj.type == MLX5_MAPPED_OBJ_SAMPLE) { + mlx5_esw_sample_skb(skb, &mapped_obj); + return false; + } +#endif /* CONFIG_MLX5_TC_SAMPLE */ + if (mapped_obj.type != MLX5_MAPPED_OBJ_SAMPLE && + mapped_obj.type != MLX5_MAPPED_OBJ_CHAIN) { netdev_dbg(priv->netdev, "Invalid mapped object type: %d\n", mapped_obj.type); return false; } -#endif /* CONFIG_NET_TC_SKB_EXT */ return true; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c index 79a0e49b2799..629a0a28a7ba 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c @@ -298,6 +298,21 @@ sample_restore_put(struct mlx5_esw_psample *esw_psample, struct mlx5_sample_rest } } +void mlx5_esw_sample_skb(struct sk_buff *skb, struct mlx5_mapped_obj *mapped_obj) +{ + u32 trunc_size = mapped_obj->sample.trunc_size; + struct psample_group psample_group = {}; + struct psample_metadata md = {}; + + md.trunc_size = trunc_size ? min(trunc_size, skb->len) : skb->len; + md.in_ifindex = skb->dev->ifindex; + psample_group.group_num = mapped_obj->sample.group_id; + psample_group.net = &init_net; + skb_push(skb, skb->mac_len); + + psample_sample_packet(&psample_group, skb, mapped_obj->sample.rate, &md); +} + struct mlx5_esw_psample * mlx5_esw_sample_init(struct mlx5e_priv *priv) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h index e42e3cb01c8c..82bff97bd9b7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h @@ -5,6 +5,7 @@ #define __MLX5_EN_TC_SAMPLE_H__ #include "en.h" +#include "eswitch.h" struct mlx5_sample_attr { u32 group_num; @@ -12,6 +13,8 @@ struct mlx5_sample_attr { u32 trunc_size; }; +void mlx5_esw_sample_skb(struct sk_buff *skb, struct mlx5_mapped_obj *mapped_obj); + struct mlx5_esw_psample * mlx5_esw_sample_init(struct mlx5e_priv *priv); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 9b26bd67e2b8..ed4d7f8f798f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -50,12 +50,18 @@ enum mlx5_mapped_obj_type { MLX5_MAPPED_OBJ_CHAIN, + MLX5_MAPPED_OBJ_SAMPLE, }; struct mlx5_mapped_obj { enum mlx5_mapped_obj_type type; union { u32 chain; + struct { + u32 group_id; + u32 rate; + u32 trunc_size; + } sample; }; }; From f94d6389f6a8abd04760dcd015d14961260a8000 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Mon, 21 Sep 2020 16:45:07 +0800 Subject: [PATCH 13/13] net/mlx5e: TC, Add support to offload sample action The following diagram illustrates the hardware model for tc sample action: +---------------------+ + original flow table + +---------------------+ + original match + +---------------------+ | v +------------------------------------------------+ + Flow Sampler Object + +------------------------------------------------+ + sample ratio + +------------------------------------------------+ + sample table id | default table id + +------------------------------------------------+ | | v v +-----------------------------+ +----------------------------------------+ + sample table + + default table per + +-----------------------------+ +----------------------------------------+ + forward to management vport + + original match + +-----------------------------+ +----------------------------------------+ + other actions + +----------------------------------------+ The sample action is translated to a goto flow table object destination which samples packets according to the provided sample ratio. Sampled packets are duplicated. One copy is processed by a termination table, named the sample table, which sends the packet to the eswitch manager port (that will be processed by software). The second copy is processed by the default table which executes the subsequent actions. The default table is created per tuple as rules with different prios and chains may overlap. Signed-off-by: Chris Mi Reviewed-by: Oz Shlomo Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en_tc.c | 30 +++ .../ethernet/mellanox/mlx5/core/esw/sample.c | 235 ++++++++++++++++++ .../ethernet/mellanox/mlx5/core/esw/sample.h | 18 ++ .../net/ethernet/mellanox/mlx5/core/eswitch.h | 4 + .../mellanox/mlx5/core/eswitch_offloads.c | 43 +++- 5 files changed, 328 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 1a403112defd..d157d1b9cad6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -223,6 +223,25 @@ get_ct_priv(struct mlx5e_priv *priv) return priv->fs.tc.ct; } +#if IS_ENABLED(CONFIG_MLX5_TC_SAMPLE) +static struct mlx5_esw_psample * +get_sample_priv(struct mlx5e_priv *priv) +{ + struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; + struct mlx5_rep_uplink_priv *uplink_priv; + struct mlx5e_rep_priv *uplink_rpriv; + + if (is_mdev_switchdev_mode(priv->mdev)) { + uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); + uplink_priv = &uplink_rpriv->uplink_priv; + + return uplink_priv->esw_psample; + } + + return NULL; +} +#endif + struct mlx5_flow_handle * mlx5_tc_rule_insert(struct mlx5e_priv *priv, struct mlx5_flow_spec *spec, @@ -1092,6 +1111,10 @@ mlx5e_tc_offload_fdb_rules(struct mlx5_eswitch *esw, rule = mlx5_tc_ct_flow_offload(get_ct_priv(flow->priv), flow, spec, attr, mod_hdr_acts); +#if IS_ENABLED(CONFIG_MLX5_TC_SAMPLE) + } else if (flow_flag_test(flow, SAMPLE)) { + rule = mlx5_esw_sample_offload(get_sample_priv(flow->priv), spec, attr); +#endif } else { rule = mlx5_eswitch_add_offloaded_rule(esw, spec, attr); } @@ -1127,6 +1150,13 @@ void mlx5e_tc_unoffload_fdb_rules(struct mlx5_eswitch *esw, return; } +#if IS_ENABLED(CONFIG_MLX5_TC_SAMPLE) + if (flow_flag_test(flow, SAMPLE)) { + mlx5_esw_sample_unoffload(get_sample_priv(flow->priv), flow->rule[0], attr); + return; + } +#endif + if (attr->esw_attr->split_count) mlx5_eswitch_del_fwd_rule(esw, flow->rule[1], attr); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c index 629a0a28a7ba..794012c5c476 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.c @@ -3,11 +3,20 @@ #include #include +#include "en/mapping.h" #include "esw/sample.h" #include "eswitch.h" #include "en_tc.h" #include "fs_core.h" +#define MLX5_ESW_VPORT_TBL_SIZE_SAMPLE (64 * 1024) + +static const struct esw_vport_tbl_namespace mlx5_esw_vport_tbl_sample_ns = { + .max_fte = MLX5_ESW_VPORT_TBL_SIZE_SAMPLE, + .max_num_groups = 0, /* default num of groups */ + .flags = MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT | MLX5_FLOW_TABLE_TUNNEL_EN_DECAP, +}; + struct mlx5_esw_psample { struct mlx5e_priv *priv; struct mlx5_flow_table *termtbl; @@ -30,6 +39,9 @@ struct mlx5_sampler { struct mlx5_sample_flow { struct mlx5_sampler *sampler; struct mlx5_sample_restore *restore; + struct mlx5_flow_attr *pre_attr; + struct mlx5_flow_handle *pre_rule; + struct mlx5_flow_handle *rule; }; struct mlx5_sample_restore { @@ -313,6 +325,229 @@ void mlx5_esw_sample_skb(struct sk_buff *skb, struct mlx5_mapped_obj *mapped_obj psample_sample_packet(&psample_group, skb, mapped_obj->sample.rate, &md); } +/* For the following typical flow table: + * + * +-------------------------------+ + * + original flow table + + * +-------------------------------+ + * + original match + + * +-------------------------------+ + * + sample action + other actions + + * +-------------------------------+ + * + * We translate the tc filter with sample action to the following HW model: + * + * +---------------------+ + * + original flow table + + * +---------------------+ + * + original match + + * +---------------------+ + * | + * v + * +------------------------------------------------+ + * + Flow Sampler Object + + * +------------------------------------------------+ + * + sample ratio + + * +------------------------------------------------+ + * + sample table id | default table id + + * +------------------------------------------------+ + * | | + * v v + * +-----------------------------+ +----------------------------------------+ + * + sample table + + default table per + + * +-----------------------------+ +----------------------------------------+ + * + forward to management vport + + original match + + * +-----------------------------+ +----------------------------------------+ + * + other actions + + * +----------------------------------------+ + */ +struct mlx5_flow_handle * +mlx5_esw_sample_offload(struct mlx5_esw_psample *esw_psample, + struct mlx5_flow_spec *spec, + struct mlx5_flow_attr *attr) +{ + struct mlx5_esw_flow_attr *esw_attr = attr->esw_attr; + struct mlx5_vport_tbl_attr per_vport_tbl_attr; + struct mlx5_esw_flow_attr *pre_esw_attr; + struct mlx5_mapped_obj restore_obj = {}; + struct mlx5_sample_flow *sample_flow; + struct mlx5_sample_attr *sample_attr; + struct mlx5_flow_table *default_tbl; + struct mlx5_flow_attr *pre_attr; + struct mlx5_eswitch *esw; + u32 obj_id; + int err; + + if (IS_ERR_OR_NULL(esw_psample)) + return ERR_PTR(-EOPNOTSUPP); + + /* If slow path flag is set, eg. when the neigh is invalid for encap, + * don't offload sample action. + */ + esw = esw_psample->priv->mdev->priv.eswitch; + if (attr->flags & MLX5_ESW_ATTR_FLAG_SLOW_PATH) + return mlx5_eswitch_add_offloaded_rule(esw, spec, attr); + + sample_flow = kzalloc(sizeof(*sample_flow), GFP_KERNEL); + if (!sample_flow) + return ERR_PTR(-ENOMEM); + esw_attr->sample->sample_flow = sample_flow; + + /* Allocate default table per vport, chain and prio. Otherwise, there is + * only one default table for the same sampler object. Rules with different + * prio and chain may overlap. For CT sample action, per vport default + * table is needed to resotre the metadata. + */ + per_vport_tbl_attr.chain = attr->chain; + per_vport_tbl_attr.prio = attr->prio; + per_vport_tbl_attr.vport = esw_attr->in_rep->vport; + per_vport_tbl_attr.vport_ns = &mlx5_esw_vport_tbl_sample_ns; + default_tbl = mlx5_esw_vporttbl_get(esw, &per_vport_tbl_attr); + if (IS_ERR(default_tbl)) { + err = PTR_ERR(default_tbl); + goto err_default_tbl; + } + + /* Perform the original matches on the default table. + * Offload all actions except the sample action. + */ + esw_attr->sample->sample_default_tbl = default_tbl; + /* When offloading sample and encap action, if there is no valid + * neigh data struct, a slow path rule is offloaded first. Source + * port metadata match is set at that time. A per vport table is + * already allocated. No need to match it again. So clear the source + * port metadata match. + */ + mlx5_eswitch_clear_rule_source_port(esw, spec); + sample_flow->rule = mlx5_eswitch_add_offloaded_rule(esw, spec, attr); + if (IS_ERR(sample_flow->rule)) { + err = PTR_ERR(sample_flow->rule); + goto err_offload_rule; + } + + /* Create sampler object. */ + sample_flow->sampler = sampler_get(esw_psample, esw_attr->sample->rate, default_tbl->id); + if (IS_ERR(sample_flow->sampler)) { + err = PTR_ERR(sample_flow->sampler); + goto err_sampler; + } + + /* Create an id mapping reg_c0 value to sample object. */ + restore_obj.type = MLX5_MAPPED_OBJ_SAMPLE; + restore_obj.sample.group_id = esw_attr->sample->group_num; + restore_obj.sample.rate = esw_attr->sample->rate; + restore_obj.sample.trunc_size = esw_attr->sample->trunc_size; + err = mapping_add(esw->offloads.reg_c0_obj_pool, &restore_obj, &obj_id); + if (err) + goto err_obj_id; + esw_attr->sample->restore_obj_id = obj_id; + + /* Create sample restore context. */ + sample_flow->restore = sample_restore_get(esw_psample, obj_id); + if (IS_ERR(sample_flow->restore)) { + err = PTR_ERR(sample_flow->restore); + goto err_sample_restore; + } + + /* Perform the original matches on the original table. Offload the + * sample action. The destination is the sampler object. + */ + pre_attr = mlx5_alloc_flow_attr(MLX5_FLOW_NAMESPACE_FDB); + if (!pre_attr) { + err = -ENOMEM; + goto err_alloc_flow_attr; + } + sample_attr = kzalloc(sizeof(*sample_attr), GFP_KERNEL); + if (!sample_attr) { + err = -ENOMEM; + goto err_alloc_sample_attr; + } + pre_esw_attr = pre_attr->esw_attr; + pre_attr->action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | MLX5_FLOW_CONTEXT_ACTION_MOD_HDR; + pre_attr->modify_hdr = sample_flow->restore->modify_hdr; + pre_attr->flags = MLX5_ESW_ATTR_FLAG_SAMPLE; + pre_attr->chain = attr->chain; + pre_attr->prio = attr->prio; + pre_esw_attr->sample = sample_attr; + pre_esw_attr->sample->sampler_id = sample_flow->sampler->sampler_id; + pre_esw_attr->in_mdev = esw_attr->in_mdev; + pre_esw_attr->in_rep = esw_attr->in_rep; + sample_flow->pre_rule = mlx5_eswitch_add_offloaded_rule(esw, spec, pre_attr); + if (IS_ERR(sample_flow->pre_rule)) { + err = PTR_ERR(sample_flow->pre_rule); + goto err_pre_offload_rule; + } + sample_flow->pre_attr = pre_attr; + + return sample_flow->rule; + +err_pre_offload_rule: + kfree(sample_attr); +err_alloc_sample_attr: + kfree(pre_attr); +err_alloc_flow_attr: + sample_restore_put(esw_psample, sample_flow->restore); +err_sample_restore: + mapping_remove(esw->offloads.reg_c0_obj_pool, obj_id); +err_obj_id: + sampler_put(esw_psample, sample_flow->sampler); +err_sampler: + /* For sample offload, rule is added in default_tbl. No need to call + * mlx5_esw_chains_put_table() + */ + attr->prio = 0; + attr->chain = 0; + mlx5_eswitch_del_offloaded_rule(esw, sample_flow->rule, attr); +err_offload_rule: + mlx5_esw_vporttbl_put(esw, &per_vport_tbl_attr); +err_default_tbl: + return ERR_PTR(err); +} + +void +mlx5_esw_sample_unoffload(struct mlx5_esw_psample *esw_psample, + struct mlx5_flow_handle *rule, + struct mlx5_flow_attr *attr) +{ + struct mlx5_esw_flow_attr *esw_attr = attr->esw_attr; + struct mlx5_sample_flow *sample_flow; + struct mlx5_vport_tbl_attr tbl_attr; + struct mlx5_flow_attr *pre_attr; + struct mlx5_eswitch *esw; + + if (IS_ERR_OR_NULL(esw_psample)) + return; + + /* If slow path flag is set, sample action is not offloaded. + * No need to delete sample rule. + */ + esw = esw_psample->priv->mdev->priv.eswitch; + if (attr->flags & MLX5_ESW_ATTR_FLAG_SLOW_PATH) { + mlx5_eswitch_del_offloaded_rule(esw, rule, attr); + return; + } + + sample_flow = esw_attr->sample->sample_flow; + pre_attr = sample_flow->pre_attr; + memset(pre_attr, 0, sizeof(*pre_attr)); + esw = esw_psample->priv->mdev->priv.eswitch; + mlx5_eswitch_del_offloaded_rule(esw, sample_flow->pre_rule, pre_attr); + mlx5_eswitch_del_offloaded_rule(esw, sample_flow->rule, attr); + + sample_restore_put(esw_psample, sample_flow->restore); + mapping_remove(esw->offloads.reg_c0_obj_pool, esw_attr->sample->restore_obj_id); + sampler_put(esw_psample, sample_flow->sampler); + tbl_attr.chain = attr->chain; + tbl_attr.prio = attr->prio; + tbl_attr.vport = esw_attr->in_rep->vport; + tbl_attr.vport_ns = &mlx5_esw_vport_tbl_sample_ns; + mlx5_esw_vporttbl_put(esw, &tbl_attr); + + kfree(pre_attr->esw_attr->sample); + kfree(pre_attr); + kfree(sample_flow); +} + struct mlx5_esw_psample * mlx5_esw_sample_init(struct mlx5e_priv *priv) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h index 82bff97bd9b7..2a3f4be10030 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/sample.h @@ -7,14 +7,32 @@ #include "en.h" #include "eswitch.h" +struct mlx5e_priv; +struct mlx5_flow_attr; +struct mlx5_esw_psample; + struct mlx5_sample_attr { u32 group_num; u32 rate; u32 trunc_size; + u32 restore_obj_id; + u32 sampler_id; + struct mlx5_flow_table *sample_default_tbl; + struct mlx5_sample_flow *sample_flow; }; void mlx5_esw_sample_skb(struct sk_buff *skb, struct mlx5_mapped_obj *mapped_obj); +struct mlx5_flow_handle * +mlx5_esw_sample_offload(struct mlx5_esw_psample *sample_priv, + struct mlx5_flow_spec *spec, + struct mlx5_flow_attr *attr); + +void +mlx5_esw_sample_unoffload(struct mlx5_esw_psample *sample_priv, + struct mlx5_flow_handle *rule, + struct mlx5_flow_attr *attr); + struct mlx5_esw_psample * mlx5_esw_sample_init(struct mlx5e_priv *priv); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index ed4d7f8f798f..deafb0e03787 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -376,6 +376,9 @@ void mlx5_eswitch_termtbl_put(struct mlx5_eswitch *esw, struct mlx5_termtbl_handle *tt); +void +mlx5_eswitch_clear_rule_source_port(struct mlx5_eswitch *esw, struct mlx5_flow_spec *spec); + struct mlx5_flow_handle * mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, struct mlx5_flow_spec *spec, @@ -423,6 +426,7 @@ enum { MLX5_ESW_ATTR_FLAG_SLOW_PATH = BIT(1), MLX5_ESW_ATTR_FLAG_NO_IN_PORT = BIT(2), MLX5_ESW_ATTR_FLAG_SRC_REWRITE = BIT(3), + MLX5_ESW_ATTR_FLAG_SAMPLE = BIT(4), }; struct mlx5_esw_flow_attr { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 510e9b8b24fe..ac92ffc8a5d3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -86,6 +86,26 @@ mlx5_eswitch_set_rule_flow_source(struct mlx5_eswitch *esw, MLX5_FLOW_CONTEXT_FLOW_SOURCE_LOCAL_VPORT; } +/* Actually only the upper 16 bits of reg c0 need to be cleared, but the lower 16 bits + * are not needed as well in the following process. So clear them all for simplicity. + */ +void +mlx5_eswitch_clear_rule_source_port(struct mlx5_eswitch *esw, struct mlx5_flow_spec *spec) +{ + if (mlx5_eswitch_vport_match_metadata_enabled(esw)) { + void *misc2; + + misc2 = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters_2); + MLX5_SET(fte_match_set_misc2, misc2, metadata_reg_c_0, 0); + + misc2 = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters_2); + MLX5_SET(fte_match_set_misc2, misc2, metadata_reg_c_0, 0); + + if (!memchr_inv(misc2, 0, MLX5_ST_SZ_BYTES(fte_match_set_misc2))) + spec->match_criteria_enable &= ~MLX5_MATCH_MISC_PARAMETERS_2; + } +} + static void mlx5_eswitch_set_rule_source_port(struct mlx5_eswitch *esw, struct mlx5_flow_spec *spec, @@ -156,6 +176,19 @@ esw_cleanup_decap_indir(struct mlx5_eswitch *esw, true); } +static int +esw_setup_sampler_dest(struct mlx5_flow_destination *dest, + struct mlx5_flow_act *flow_act, + struct mlx5_esw_flow_attr *esw_attr, + int i) +{ + flow_act->flags |= FLOW_ACT_IGNORE_FLOW_LEVEL; + dest[i].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_SAMPLER; + dest[i].sampler_id = esw_attr->sample->sampler_id; + + return 0; +} + static int esw_setup_ft_dest(struct mlx5_flow_destination *dest, struct mlx5_flow_act *flow_act, @@ -385,7 +418,10 @@ esw_setup_dests(struct mlx5_flow_destination *dest, MLX5_CAP_ESW_FLOWTABLE_FDB(esw->dev, ignore_flow_level)) attr->flags |= MLX5_ESW_ATTR_FLAG_SRC_REWRITE; - if (attr->dest_ft) { + if (attr->flags & MLX5_ESW_ATTR_FLAG_SAMPLE) { + esw_setup_sampler_dest(dest, flow_act, esw_attr, *i); + (*i)++; + } else if (attr->dest_ft) { esw_setup_ft_dest(dest, flow_act, esw, attr, spec, *i); (*i)++; } else if (attr->flags & MLX5_ESW_ATTR_FLAG_SLOW_PATH) { @@ -488,7 +524,10 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) flow_act.modify_hdr = attr->modify_hdr; - if (split) { + /* esw_attr->sample is allocated only when there is a sample action */ + if (esw_attr->sample && esw_attr->sample->sample_default_tbl) { + fdb = esw_attr->sample->sample_default_tbl; + } else if (split) { fwd_attr.chain = attr->chain; fwd_attr.prio = attr->prio; fwd_attr.vport = esw_attr->in_rep->vport;