From: Carolina Jubran cjubran@nvidia.com

Introduce support for specifying relative bandwidth shares between traffic classes (TC) in the devlink-rate API. This new option allows users to allocate bandwidth across multiple traffic classes in a single command.

This feature provides a more granular control over traffic management, especially for scenarios requiring Enhanced Transmission Selection.

Users can now define a relative bandwidth share for each traffic class. For example, assigning share values of 20 to TC0 (TCP/UDP) and 80 to TC5 (RoCE) will result in TC0 receiving 20% and TC5 receiving 80% of the total bandwidth. The actual percentage each class receives depends on the ratio of its share value to the sum of all shares.

Example: DEV=pci/0000:08:00.0

$ devlink port function rate add $DEV/vfs_group tx_share 10Gbit \ tx_max 50Gbit tc-bw 0:20 1:0 2:0 3:0 4:0 5:80 6:0 7:0

$ devlink port function rate set $DEV/vfs_group \ tc-bw 0:20 1:0 2:0 3:0 4:0 5:20 6:60 7:0

Example usage with ynl:

./tools/net/ynl/cli.py --spec Documentation/netlink/specs/devlink.yaml \ --do rate-set --json '{ "bus-name": "pci", "dev-name": "0000:08:00.0", "port-index": 1, "rate-tc-bws": [ {"rate-tc-index": 0, "rate-tc-bw": 50}, {"rate-tc-index": 1, "rate-tc-bw": 50}, {"rate-tc-index": 2, "rate-tc-bw": 0}, {"rate-tc-index": 3, "rate-tc-bw": 0}, {"rate-tc-index": 4, "rate-tc-bw": 0}, {"rate-tc-index": 5, "rate-tc-bw": 0}, {"rate-tc-index": 6, "rate-tc-bw": 0}, {"rate-tc-index": 7, "rate-tc-bw": 0} ] }'

./tools/net/ynl/cli.py --spec Documentation/netlink/specs/devlink.yaml \ --do rate-get --json '{ "bus-name": "pci", "dev-name": "0000:08:00.0", "port-index": 1 }'

output for rate-get: {'bus-name': 'pci', 'dev-name': '0000:08:00.0', 'port-index': 1, 'rate-tc-bws': [{'rate-tc-bw': 50, 'rate-tc-index': 0}, {'rate-tc-bw': 50, 'rate-tc-index': 1}, {'rate-tc-bw': 0, 'rate-tc-index': 2}, {'rate-tc-bw': 0, 'rate-tc-index': 3}, {'rate-tc-bw': 0, 'rate-tc-index': 4}, {'rate-tc-bw': 0, 'rate-tc-index': 5}, {'rate-tc-bw': 0, 'rate-tc-index': 6}, {'rate-tc-bw': 0, 'rate-tc-index': 7}], 'rate-tx-max': 0, 'rate-tx-priority': 0, 'rate-tx-share': 0, 'rate-tx-weight': 0, 'rate-type': 'leaf'}

Signed-off-by: Carolina Jubran cjubran@nvidia.com Reviewed-by: Cosmin Ratiu cratiu@nvidia.com Reviewed-by: Jiri Pirko jiri@nvidia.com Signed-off-by: Tariq Toukan tariqt@nvidia.com Signed-off-by: Mark Bloch mbloch@nvidia.com --- Documentation/netlink/specs/devlink.yaml | 32 ++++- .../networking/devlink/devlink-port.rst | 8 ++ include/net/devlink.h | 8 ++ include/uapi/linux/devlink.h | 9 ++ net/devlink/netlink_gen.c | 15 +- net/devlink/netlink_gen.h | 1 + net/devlink/rate.c | 129 ++++++++++++++++++ 7 files changed, 197 insertions(+), 5 deletions(-)

diff --git a/Documentation/netlink/specs/devlink.yaml b/Documentation/netlink/specs/devlink.yaml index ad56093d6ead..c66ea7e325b7 100644 --- a/Documentation/netlink/specs/devlink.yaml +++ b/Documentation/netlink/specs/devlink.yaml @@ -224,6 +224,10 @@ definitions: value: 10 - name: binary + - + name: rate-tc-index-max + type: const + value: 7

attribute-sets: - @@ -844,7 +848,23 @@ attribute-sets: - name: region-direct type: flag - + - + name: rate-tc-bws + type: nest + multi-attr: true + nested-attributes: dl-rate-tc-bws + - + name: rate-tc-index + type: u8 + checks: + max: rate-tc-index-max + - + name: rate-tc-bw + type: u32 + doc: | + Specifies the bandwidth share assigned to the Traffic Class. + The bandwidth for the traffic class is determined + in proportion to the sum of the shares of all configured classes. - name: dl-dev-stats subset-of: devlink @@ -1249,6 +1269,14 @@ attribute-sets: - name: flash type: flag + - + name: dl-rate-tc-bws + subset-of: devlink + attributes: + - + name: rate-tc-index + - + name: rate-tc-bw

operations: enum-model: directional @@ -2176,6 +2204,7 @@ operations: - rate-tx-priority - rate-tx-weight - rate-parent-node-name + - rate-tc-bws

- name: rate-new @@ -2196,6 +2225,7 @@ operations: - rate-tx-priority - rate-tx-weight - rate-parent-node-name + - rate-tc-bws

- name: rate-del diff --git a/Documentation/networking/devlink/devlink-port.rst b/Documentation/networking/devlink/devlink-port.rst index 9d22d41a7cd1..5e397798a402 100644 --- a/Documentation/networking/devlink/devlink-port.rst +++ b/Documentation/networking/devlink/devlink-port.rst @@ -418,6 +418,14 @@ API allows to configure following rate object's parameters: to all node children limits. ``tx_max`` is an upper limit for children. ``tx_share`` is a total bandwidth distributed among children.

+``tc_bw`` + Allow users to set the bandwidth allocation per traffic class on rate + objects. This enables fine-grained QoS configurations by assigning a relative + share value to each traffic class. The bandwidth is distributed in proportion + to the share value for each class, relative to the sum of all shares. + When applied to a non-leaf node, tc_bw determines how bandwidth is shared + among its child elements. + ``tx_priority`` and ``tx_weight`` can be used simultaneously. In that case nodes with the same priority form a WFQ subgroup in the sibling group and arbitration among them is based on assigned weights. diff --git a/include/net/devlink.h b/include/net/devlink.h index 63517646a497..d0ce5a7e984c 100644 --- a/include/net/devlink.h +++ b/include/net/devlink.h @@ -118,6 +118,8 @@ struct devlink_rate {

u32 tx_priority; u32 tx_weight; + + u32 tc_bw[DEVLINK_RATE_TCS_MAX]; };

struct devlink_port { @@ -1486,6 +1488,9 @@ struct devlink_ops { u32 tx_priority, struct netlink_ext_ack *extack); int (*rate_leaf_tx_weight_set)(struct devlink_rate *devlink_rate, void *priv, u32 tx_weight, struct netlink_ext_ack *extack); + int (*rate_leaf_tc_bw_set)(struct devlink_rate *devlink_rate, + void *priv, u32 *tc_bw, + struct netlink_ext_ack *extack); int (*rate_node_tx_share_set)(struct devlink_rate *devlink_rate, void *priv, u64 tx_share, struct netlink_ext_ack *extack); int (*rate_node_tx_max_set)(struct devlink_rate *devlink_rate, void *priv, @@ -1494,6 +1499,9 @@ struct devlink_ops { u32 tx_priority, struct netlink_ext_ack *extack); int (*rate_node_tx_weight_set)(struct devlink_rate *devlink_rate, void *priv, u32 tx_weight, struct netlink_ext_ack *extack); + int (*rate_node_tc_bw_set)(struct devlink_rate *devlink_rate, + void *priv, u32 *tc_bw, + struct netlink_ext_ack *extack); int (*rate_node_new)(struct devlink_rate *rate_node, void **priv, struct netlink_ext_ack *extack); int (*rate_node_del)(struct devlink_rate *rate_node, void *priv, diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h index a5ee0f13740a..e72bcc239afd 100644 --- a/include/uapi/linux/devlink.h +++ b/include/uapi/linux/devlink.h @@ -221,6 +221,11 @@ enum devlink_port_flavour { */ };

+/* IEEE 802.1Qaz standard supported values. */ + +#define DEVLINK_RATE_TCS_MAX 8 +#define DEVLINK_RATE_TC_INDEX_MAX (DEVLINK_RATE_TCS_MAX - 1) + enum devlink_rate_type { DEVLINK_RATE_TYPE_LEAF, DEVLINK_RATE_TYPE_NODE, @@ -629,6 +634,10 @@ enum devlink_attr {

DEVLINK_ATTR_REGION_DIRECT, /* flag */

+ DEVLINK_ATTR_RATE_TC_BWS, /* nested */ + DEVLINK_ATTR_RATE_TC_INDEX, /* u8 */ + DEVLINK_ATTR_RATE_TC_BW, /* u32 */ + /* Add new attributes above here, update the spec in * Documentation/netlink/specs/devlink.yaml and re-generate * net/devlink/netlink_gen.c. diff --git a/net/devlink/netlink_gen.c b/net/devlink/netlink_gen.c index e340d955cf3b..c50436433c18 100644 --- a/net/devlink/netlink_gen.c +++ b/net/devlink/netlink_gen.c @@ -45,6 +45,11 @@ const struct nla_policy devlink_dl_port_function_nl_policy[DEVLINK_PORT_FN_ATTR_ [DEVLINK_PORT_FN_ATTR_CAPS] = NLA_POLICY_BITFIELD32(15), };

+const struct nla_policy devlink_dl_rate_tc_bws_nl_policy[DEVLINK_ATTR_RATE_TC_BW + 1] = { + [DEVLINK_ATTR_RATE_TC_INDEX] = NLA_POLICY_MAX(NLA_U8, DEVLINK_RATE_TC_INDEX_MAX), + [DEVLINK_ATTR_RATE_TC_BW] = { .type = NLA_U32, }, +}; + const struct nla_policy devlink_dl_selftest_id_nl_policy[DEVLINK_ATTR_SELFTEST_ID_FLASH + 1] = { [DEVLINK_ATTR_SELFTEST_ID_FLASH] = { .type = NLA_FLAG, }, }; @@ -523,7 +528,7 @@ static const struct nla_policy devlink_rate_get_dump_nl_policy[DEVLINK_ATTR_DEV_ };

/* DEVLINK_CMD_RATE_SET - do */ -static const struct nla_policy devlink_rate_set_nl_policy[DEVLINK_ATTR_RATE_TX_WEIGHT + 1] = { +static const struct nla_policy devlink_rate_set_nl_policy[DEVLINK_ATTR_RATE_TC_BWS + 1] = { [DEVLINK_ATTR_BUS_NAME] = { .type = NLA_NUL_STRING, }, [DEVLINK_ATTR_DEV_NAME] = { .type = NLA_NUL_STRING, }, [DEVLINK_ATTR_RATE_NODE_NAME] = { .type = NLA_NUL_STRING, }, @@ -532,10 +537,11 @@ static const struct nla_policy devlink_rate_set_nl_policy[DEVLINK_ATTR_RATE_TX_W [DEVLINK_ATTR_RATE_TX_PRIORITY] = { .type = NLA_U32, }, [DEVLINK_ATTR_RATE_TX_WEIGHT] = { .type = NLA_U32, }, [DEVLINK_ATTR_RATE_PARENT_NODE_NAME] = { .type = NLA_NUL_STRING, }, + [DEVLINK_ATTR_RATE_TC_BWS] = NLA_POLICY_NESTED(devlink_dl_rate_tc_bws_nl_policy), };

/* DEVLINK_CMD_RATE_NEW - do */ -static const struct nla_policy devlink_rate_new_nl_policy[DEVLINK_ATTR_RATE_TX_WEIGHT + 1] = { +static const struct nla_policy devlink_rate_new_nl_policy[DEVLINK_ATTR_RATE_TC_BWS + 1] = { [DEVLINK_ATTR_BUS_NAME] = { .type = NLA_NUL_STRING, }, [DEVLINK_ATTR_DEV_NAME] = { .type = NLA_NUL_STRING, }, [DEVLINK_ATTR_RATE_NODE_NAME] = { .type = NLA_NUL_STRING, }, @@ -544,6 +550,7 @@ static const struct nla_policy devlink_rate_new_nl_policy[DEVLINK_ATTR_RATE_TX_W [DEVLINK_ATTR_RATE_TX_PRIORITY] = { .type = NLA_U32, }, [DEVLINK_ATTR_RATE_TX_WEIGHT] = { .type = NLA_U32, }, [DEVLINK_ATTR_RATE_PARENT_NODE_NAME] = { .type = NLA_NUL_STRING, }, + [DEVLINK_ATTR_RATE_TC_BWS] = NLA_POLICY_NESTED(devlink_dl_rate_tc_bws_nl_policy), };

/* DEVLINK_CMD_RATE_DEL - do */ @@ -1191,7 +1198,7 @@ const struct genl_split_ops devlink_nl_ops[74] = { .doit = devlink_nl_rate_set_doit, .post_doit = devlink_nl_post_doit, .policy = devlink_rate_set_nl_policy, - .maxattr = DEVLINK_ATTR_RATE_TX_WEIGHT, + .maxattr = DEVLINK_ATTR_RATE_TC_BWS, .flags = GENL_ADMIN_PERM | GENL_CMD_CAP_DO, }, { @@ -1201,7 +1208,7 @@ const struct genl_split_ops devlink_nl_ops[74] = { .doit = devlink_nl_rate_new_doit, .post_doit = devlink_nl_post_doit, .policy = devlink_rate_new_nl_policy, - .maxattr = DEVLINK_ATTR_RATE_TX_WEIGHT, + .maxattr = DEVLINK_ATTR_RATE_TC_BWS, .flags = GENL_ADMIN_PERM | GENL_CMD_CAP_DO, }, { diff --git a/net/devlink/netlink_gen.h b/net/devlink/netlink_gen.h index 8f2bd50ddf5e..fb733b5d4ff1 100644 --- a/net/devlink/netlink_gen.h +++ b/net/devlink/netlink_gen.h @@ -13,6 +13,7 @@

/* Common nested types */ extern const struct nla_policy devlink_dl_port_function_nl_policy[DEVLINK_PORT_FN_ATTR_CAPS + 1]; +extern const struct nla_policy devlink_dl_rate_tc_bws_nl_policy[DEVLINK_ATTR_RATE_TC_BW + 1]; extern const struct nla_policy devlink_dl_selftest_id_nl_policy[DEVLINK_ATTR_SELFTEST_ID_FLASH + 1];

/* Ops table for devlink */ diff --git a/net/devlink/rate.c b/net/devlink/rate.c index 8828ffaf6cbc..181d443ab3fc 100644 --- a/net/devlink/rate.c +++ b/net/devlink/rate.c @@ -80,6 +80,29 @@ devlink_rate_get_from_info(struct devlink *devlink, struct genl_info *info) return ERR_PTR(-EINVAL); }

+static int devlink_rate_put_tc_bws(struct sk_buff *msg, u32 *tc_bw) +{ + struct nlattr *nla_tc_bw; + int i; + + for (i = 0; i < DEVLINK_RATE_TCS_MAX; i++) { + nla_tc_bw = nla_nest_start(msg, DEVLINK_ATTR_RATE_TC_BWS); + if (!nla_tc_bw) + return -EMSGSIZE; + + if (nla_put_u8(msg, DEVLINK_ATTR_RATE_TC_INDEX, i) || + nla_put_u32(msg, DEVLINK_ATTR_RATE_TC_BW, tc_bw[i])) + goto nla_put_failure; + + nla_nest_end(msg, nla_tc_bw); + } + return 0; + +nla_put_failure: + nla_nest_cancel(msg, nla_tc_bw); + return -EMSGSIZE; +} + static int devlink_nl_rate_fill(struct sk_buff *msg, struct devlink_rate *devlink_rate, enum devlink_command cmd, u32 portid, u32 seq, @@ -129,6 +152,9 @@ static int devlink_nl_rate_fill(struct sk_buff *msg, devlink_rate->parent->name)) goto nla_put_failure;

+ if (devlink_rate_put_tc_bws(msg, devlink_rate->tc_bw)) + goto nla_put_failure; + genlmsg_end(msg, hdr); return 0;

@@ -316,6 +342,89 @@ devlink_nl_rate_parent_node_set(struct devlink_rate *devlink_rate, return 0; }

+static int devlink_nl_rate_tc_bw_parse(struct nlattr *parent_nest, u32 *tc_bw, + unsigned long *bitmap, + struct netlink_ext_ack *extack) +{ + struct nlattr *tb[DEVLINK_ATTR_MAX + 1]; + u8 tc_index; + int err; + + err = nla_parse_nested(tb, DEVLINK_ATTR_MAX, parent_nest, + devlink_dl_rate_tc_bws_nl_policy, extack); + if (err) + return err; + + if (!tb[DEVLINK_ATTR_RATE_TC_INDEX]) { + NL_SET_ERR_ATTR_MISS(extack, parent_nest, + DEVLINK_ATTR_RATE_TC_INDEX); + return -EINVAL; + } + + tc_index = nla_get_u8(tb[DEVLINK_ATTR_RATE_TC_INDEX]); + + if (!tb[DEVLINK_ATTR_RATE_TC_BW]) { + NL_SET_ERR_ATTR_MISS(extack, parent_nest, + DEVLINK_ATTR_RATE_TC_BW); + return -EINVAL; + } + + if (test_and_set_bit(tc_index, bitmap)) { + NL_SET_ERR_MSG_FMT(extack, + "Duplicate traffic class index specified (%u)", + tc_index); + return -EINVAL; + } + + tc_bw[tc_index] = nla_get_u32(tb[DEVLINK_ATTR_RATE_TC_BW]); + + return 0; +} + +static int devlink_nl_rate_tc_bw_set(struct devlink_rate *devlink_rate, + struct genl_info *info) +{ + DECLARE_BITMAP(bitmap, DEVLINK_RATE_TCS_MAX) = {}; + struct devlink *devlink = devlink_rate->devlink; + const struct devlink_ops *ops = devlink->ops; + int rem, err = -EOPNOTSUPP, i, total = 0; + u32 tc_bw[DEVLINK_RATE_TCS_MAX] = {}; + struct nlattr *attr; + + nlmsg_for_each_attr_type(attr, DEVLINK_ATTR_RATE_TC_BWS, info->nlhdr, + GENL_HDRLEN, rem) { + err = devlink_nl_rate_tc_bw_parse(attr, tc_bw, bitmap, + info->extack); + if (err) + return err; + } + + for (i = 0; i < DEVLINK_RATE_TCS_MAX; i++) { + if (!test_bit(i, bitmap)) { + NL_SET_ERR_MSG_FMT(info->extack, + "Bandwidth values must be specified for all %u traffic classes", + DEVLINK_RATE_TCS_MAX); + return -EINVAL; + } + + total += tc_bw[i]; + } + + if (devlink_rate_is_leaf(devlink_rate)) + err = ops->rate_leaf_tc_bw_set(devlink_rate, devlink_rate->priv, + tc_bw, info->extack); + else if (devlink_rate_is_node(devlink_rate)) + err = ops->rate_node_tc_bw_set(devlink_rate, devlink_rate->priv, + tc_bw, info->extack); + + if (err) + return err; + + memcpy(devlink_rate->tc_bw, tc_bw, sizeof(tc_bw)); + + return 0; +} + static int devlink_nl_rate_set(struct devlink_rate *devlink_rate, const struct devlink_ops *ops, struct genl_info *info) @@ -388,6 +497,12 @@ static int devlink_nl_rate_set(struct devlink_rate *devlink_rate, return err; }

+ if (attrs[DEVLINK_ATTR_RATE_TC_BWS]) { + err = devlink_nl_rate_tc_bw_set(devlink_rate, info); + if (err) + return err; + } + return 0; }

@@ -423,6 +538,13 @@ static bool devlink_rate_set_ops_supported(const struct devlink_ops *ops, "TX weight set isn't supported for the leafs"); return false; } + if (attrs[DEVLINK_ATTR_RATE_TC_BWS] && + !ops->rate_leaf_tc_bw_set) { + NL_SET_ERR_MSG_ATTR(info->extack, + attrs[DEVLINK_ATTR_RATE_TC_BWS], + "TC bandwidth set isn't supported for the leafs"); + return false; + } } else if (type == DEVLINK_RATE_TYPE_NODE) { if (attrs[DEVLINK_ATTR_RATE_TX_SHARE] && !ops->rate_node_tx_share_set) { NL_SET_ERR_MSG(info->extack, "TX share set isn't supported for the nodes"); @@ -449,6 +571,13 @@ static bool devlink_rate_set_ops_supported(const struct devlink_ops *ops, "TX weight set isn't supported for the nodes"); return false; } + if (attrs[DEVLINK_ATTR_RATE_TC_BWS] && + !ops->rate_node_tc_bw_set) { + NL_SET_ERR_MSG_ATTR(info->extack, + attrs[DEVLINK_ATTR_RATE_TC_BWS], + "TC bandwidth set isn't supported for the nodes"); + return false; + } } else { WARN(1, "Unknown type of rate object"); return false;

From: Carolina Jubran cjubran@nvidia.com

Introduce `mlx5_esw_devlink_rate_node_tc_bw_set()` and `mlx5_esw_devlink_rate_leaf_tc_bw_set()` with no-op logic.

Future patches will add support for setting traffic class bandwidth on rate objects.

Signed-off-by: Carolina Jubran cjubran@nvidia.com Reviewed-by: Cosmin Ratiu cratiu@nvidia.com Signed-off-by: Tariq Toukan tariqt@nvidia.com Signed-off-by: Mark Bloch mbloch@nvidia.com --- .../net/ethernet/mellanox/mlx5/core/devlink.c | 2 ++ .../net/ethernet/mellanox/mlx5/core/esw/qos.c | 20 +++++++++++++++++++ .../net/ethernet/mellanox/mlx5/core/esw/qos.h | 8 ++++++++ 3 files changed, 30 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c index 42218834183a..3ffa3fbacd16 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c @@ -376,6 +376,8 @@ static const struct devlink_ops mlx5_devlink_ops = { .eswitch_encap_mode_get = mlx5_devlink_eswitch_encap_mode_get, .rate_leaf_tx_share_set = mlx5_esw_devlink_rate_leaf_tx_share_set, .rate_leaf_tx_max_set = mlx5_esw_devlink_rate_leaf_tx_max_set, + .rate_leaf_tc_bw_set = mlx5_esw_devlink_rate_leaf_tc_bw_set, + .rate_node_tc_bw_set = mlx5_esw_devlink_rate_node_tc_bw_set, .rate_node_tx_share_set = mlx5_esw_devlink_rate_node_tx_share_set, .rate_node_tx_max_set = mlx5_esw_devlink_rate_node_tx_max_set, .rate_node_new = mlx5_esw_devlink_rate_node_new, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c index b6ae384396b3..ec706e9352e1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c @@ -906,6 +906,26 @@ int mlx5_esw_devlink_rate_leaf_tx_max_set(struct devlink_rate *rate_leaf, void * return err; }

+int mlx5_esw_devlink_rate_leaf_tc_bw_set(struct devlink_rate *rate_leaf, + void *priv, + u32 *tc_bw, + struct netlink_ext_ack *extack) +{ + NL_SET_ERR_MSG_MOD(extack, + "TC bandwidth shares are not supported on leafs"); + return -EOPNOTSUPP; +} + +int mlx5_esw_devlink_rate_node_tc_bw_set(struct devlink_rate *rate_node, + void *priv, + u32 *tc_bw, + struct netlink_ext_ack *extack) +{ + NL_SET_ERR_MSG_MOD(extack, + "TC bandwidth shares are not supported on nodes"); + return -EOPNOTSUPP; +} + int mlx5_esw_devlink_rate_node_tx_share_set(struct devlink_rate *rate_node, void *priv, u64 tx_share, struct netlink_ext_ack *extack) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h index ed40ec8f027e..0a50982b0e27 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h @@ -21,6 +21,14 @@ int mlx5_esw_devlink_rate_leaf_tx_share_set(struct devlink_rate *rate_leaf, void u64 tx_share, struct netlink_ext_ack *extack); int mlx5_esw_devlink_rate_leaf_tx_max_set(struct devlink_rate *rate_leaf, void *priv, u64 tx_max, struct netlink_ext_ack *extack); +int mlx5_esw_devlink_rate_leaf_tc_bw_set(struct devlink_rate *rate_node, + void *priv, + u32 *tc_bw, + struct netlink_ext_ack *extack); +int mlx5_esw_devlink_rate_node_tc_bw_set(struct devlink_rate *rate_node, + void *priv, + u32 *tc_bw, + struct netlink_ext_ack *extack); int mlx5_esw_devlink_rate_node_tx_share_set(struct devlink_rate *rate_node, void *priv, u64 tx_share, struct netlink_ext_ack *extack); int mlx5_esw_devlink_rate_node_tx_max_set(struct devlink_rate *rate_node, void *priv,

From: Carolina Jubran cjubran@nvidia.com

Introduce support for traffic class (TC) scheduling on vports by allowing the vport to own multiple TC scheduling nodes. This patch enables more granular control of QoS by defining three distinct QoS states for vports, each providing unique scheduling behavior:

1. Regular QoS: The `sched_node` represents the vport directly, handling QoS as a single scheduling entity. 2. TC QoS on the vport: The `sched_node` acts as a TC arbiter, enabling TC scheduling directly on the vport. 3. TC QoS on the parent node: The `sched_node` functions as a rate limiter, with TC arbitration enabled at the parent level, associating multiple scheduling nodes with each vport.

Key changes include:

- Added support for new scheduling elements, vport traffic class and rate limiter.

- New helper functions for creating, destroying, and restoring vport TC scheduling nodes, handling transitions between regular QoS and TC arbitration states.

- Updated `esw_qos_vport_enable()` and `esw_qos_vport_disable()` to support both regular QoS and TC arbitration states, ensuring consistent transitions between scheduling modes.

- Introduced a `sched_nodes` array under `vport->qos` to store multiple TC scheduling nodes per vport, enabling finer control over per-TC QoS.

- Enhanced `esw_qos_vport_update_parent()` to handle transitions between the three QoS states based on the current and new parent node types.

This patch lays the groundwork for future support for configuring tc-bw on vports. Although the infrastructure is in place, full support for tc-bw is not yet implemented; attempts to set tc-bw on vports will return `-EOPNOTSUPP`.

No functional changes are introduced at this stage.

Signed-off-by: Carolina Jubran cjubran@nvidia.com Reviewed-by: Cosmin Ratiu cratiu@nvidia.com Signed-off-by: Tariq Toukan tariqt@nvidia.com Signed-off-by: Mark Bloch mbloch@nvidia.com --- .../net/ethernet/mellanox/mlx5/core/esw/qos.c | 438 ++++++++++++++++-- .../net/ethernet/mellanox/mlx5/core/eswitch.h | 14 +- 2 files changed, 422 insertions(+), 30 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c index 1066992c1503..dec3bed682b7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c @@ -65,12 +65,16 @@ enum sched_node_type { SCHED_NODE_TYPE_VPORTS_TSAR, SCHED_NODE_TYPE_VPORT, SCHED_NODE_TYPE_TC_ARBITER_TSAR, + SCHED_NODE_TYPE_RATE_LIMITER, + SCHED_NODE_TYPE_VPORT_TC, };

static const char * const sched_node_type_str[] = { [SCHED_NODE_TYPE_VPORTS_TSAR] = "vports TSAR", [SCHED_NODE_TYPE_VPORT] = "vport", [SCHED_NODE_TYPE_TC_ARBITER_TSAR] = "TC Arbiter TSAR", + [SCHED_NODE_TYPE_RATE_LIMITER] = "Rate Limiter", + [SCHED_NODE_TYPE_VPORT_TC] = "vport TC", };

struct mlx5_esw_sched_node { @@ -94,6 +98,8 @@ struct mlx5_esw_sched_node { struct mlx5_vport *vport; /* Level in the hierarchy. The root node level is 1. */ u8 level; + /* Valid only when this node represents a traffic class. */ + u8 tc; };

static void esw_qos_node_attach_to_parent(struct mlx5_esw_sched_node *node) @@ -148,6 +154,15 @@ static void esw_qos_nodes_set_parent(struct list_head *nodes,

void mlx5_esw_qos_vport_qos_free(struct mlx5_vport *vport) { + if (vport->qos.sched_nodes) { + int num_tcs = esw_qos_num_tcs(vport->qos.sched_node->esw->dev); + int i; + + for (i = 0; i < num_tcs; i++) + kfree(vport->qos.sched_nodes[i]); + kfree(vport->qos.sched_nodes); + } + kfree(vport->qos.sched_node); memset(&vport->qos, 0, sizeof(vport->qos)); } @@ -172,11 +187,19 @@ mlx5_esw_qos_vport_get_parent(const struct mlx5_vport *vport) static void esw_qos_sched_elem_warn(struct mlx5_esw_sched_node *node, int err, const char *op) { switch (node->type) { + case SCHED_NODE_TYPE_VPORT_TC: + esw_warn(node->esw->dev, + "E-Switch %s %s scheduling element failed (vport=%d,tc=%d,err=%d)\n", + op, + sched_node_type_str[node->type], + node->vport->vport, node->tc, err); + break; case SCHED_NODE_TYPE_VPORT: esw_warn(node->esw->dev, "E-Switch %s %s scheduling element failed (vport=%d,err=%d)\n", op, sched_node_type_str[node->type], node->vport->vport, err); break; + case SCHED_NODE_TYPE_RATE_LIMITER: case SCHED_NODE_TYPE_TC_ARBITER_TSAR: case SCHED_NODE_TYPE_VPORTS_TSAR: esw_warn(node->esw->dev, @@ -271,6 +294,24 @@ static int esw_qos_sched_elem_config(struct mlx5_esw_sched_node *node, u32 max_r return 0; }

+static int esw_qos_create_rate_limit_element(struct mlx5_esw_sched_node *node, + struct netlink_ext_ack *extack) +{ + u32 sched_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; + + if (!mlx5_qos_element_type_supported( + node->esw->dev, + SCHEDULING_CONTEXT_ELEMENT_TYPE_RATE_LIMIT, + SCHEDULING_HIERARCHY_E_SWITCH)) + return -EOPNOTSUPP; + + MLX5_SET(scheduling_context, sched_ctx, max_average_bw, node->max_rate); + MLX5_SET(scheduling_context, sched_ctx, element_type, + SCHEDULING_CONTEXT_ELEMENT_TYPE_RATE_LIMIT); + + return esw_qos_node_create_sched_element(node, sched_ctx, extack); +} + static u32 esw_qos_calculate_min_rate_divider(struct mlx5_eswitch *esw, struct mlx5_esw_sched_node *parent) { @@ -388,28 +429,64 @@ esw_qos_create_node_sched_elem(struct mlx5_core_dev *dev, u32 parent_element_id, tsar_ix); }

-static int esw_qos_vport_create_sched_element(struct mlx5_esw_sched_node *vport_node, - struct netlink_ext_ack *extack) +static int +esw_qos_vport_create_sched_element(struct mlx5_esw_sched_node *vport_node, + struct netlink_ext_ack *extack) { u32 sched_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; struct mlx5_core_dev *dev = vport_node->esw->dev; void *attr;

- if (!mlx5_qos_element_type_supported(dev, - SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT, - SCHEDULING_HIERARCHY_E_SWITCH)) + if (!mlx5_qos_element_type_supported( + dev, + SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT, + SCHEDULING_HIERARCHY_E_SWITCH)) return -EOPNOTSUPP;

MLX5_SET(scheduling_context, sched_ctx, element_type, SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT); attr = MLX5_ADDR_OF(scheduling_context, sched_ctx, element_attributes); MLX5_SET(vport_element, attr, vport_number, vport_node->vport->vport); - MLX5_SET(scheduling_context, sched_ctx, parent_element_id, vport_node->parent->ix); - MLX5_SET(scheduling_context, sched_ctx, max_average_bw, vport_node->max_rate); + MLX5_SET(scheduling_context, sched_ctx, parent_element_id, + vport_node->parent->ix); + MLX5_SET(scheduling_context, sched_ctx, max_average_bw, + vport_node->max_rate);

return esw_qos_node_create_sched_element(vport_node, sched_ctx, extack); }

+static int +esw_qos_vport_tc_create_sched_element(struct mlx5_esw_sched_node *vport_tc_node, + u32 rate_limit_elem_ix, + struct netlink_ext_ack *extack) +{ + u32 sched_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; + struct mlx5_core_dev *dev = vport_tc_node->esw->dev; + void *attr; + + if (!mlx5_qos_element_type_supported( + dev, + SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT_TC, + SCHEDULING_HIERARCHY_E_SWITCH)) + return -EOPNOTSUPP; + + MLX5_SET(scheduling_context, sched_ctx, element_type, + SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT_TC); + attr = MLX5_ADDR_OF(scheduling_context, sched_ctx, element_attributes); + MLX5_SET(vport_tc_element, attr, vport_number, + vport_tc_node->vport->vport); + MLX5_SET(vport_tc_element, attr, traffic_class, vport_tc_node->tc); + MLX5_SET(scheduling_context, sched_ctx, max_bw_obj_id, + rate_limit_elem_ix); + MLX5_SET(scheduling_context, sched_ctx, parent_element_id, + vport_tc_node->parent->ix); + MLX5_SET(scheduling_context, sched_ctx, bw_share, + vport_tc_node->bw_share); + + return esw_qos_node_create_sched_element(vport_tc_node, sched_ctx, + extack); +} + static struct mlx5_esw_sched_node * __esw_qos_alloc_node(struct mlx5_eswitch *esw, u32 tsar_ix, enum sched_node_type type, struct mlx5_esw_sched_node *parent) @@ -617,12 +694,202 @@ static int esw_qos_tc_arbiter_scheduling_setup(struct mlx5_esw_sched_node *node, return -EOPNOTSUPP; }

+static int +esw_qos_create_vport_tc_sched_node(struct mlx5_vport *vport, + u32 rate_limit_elem_ix, + struct mlx5_esw_sched_node *vports_tc_node, + struct netlink_ext_ack *extack) +{ + struct mlx5_esw_sched_node *vport_node = vport->qos.sched_node; + struct mlx5_esw_sched_node *vport_tc_node; + u8 tc = vports_tc_node->tc; + int err; + + vport_tc_node = __esw_qos_alloc_node(vport_node->esw, 0, + SCHED_NODE_TYPE_VPORT_TC, + vports_tc_node); + if (!vport_tc_node) + return -ENOMEM; + + vport_tc_node->min_rate = vport_node->min_rate; + vport_tc_node->tc = tc; + vport_tc_node->vport = vport; + err = esw_qos_vport_tc_create_sched_element(vport_tc_node, + rate_limit_elem_ix, + extack); + if (err) + goto err_out; + + vport->qos.sched_nodes[tc] = vport_tc_node; + + return 0; +err_out: + __esw_qos_free_node(vport_tc_node); + return err; +} + +static void +esw_qos_destroy_vport_tc_sched_elements(struct mlx5_vport *vport, + struct netlink_ext_ack *extack) +{ + int i, num_tcs = esw_qos_num_tcs(vport->qos.sched_node->esw->dev); + + for (i = 0; i < num_tcs; i++) { + if (vport->qos.sched_nodes[i]) { + __esw_qos_destroy_node(vport->qos.sched_nodes[i], + extack); + } + } + + kfree(vport->qos.sched_nodes); + vport->qos.sched_nodes = NULL; +} + +static int +esw_qos_create_vport_tc_sched_elements(struct mlx5_vport *vport, + enum sched_node_type type, + struct netlink_ext_ack *extack) +{ + struct mlx5_esw_sched_node *vport_node = vport->qos.sched_node; + struct mlx5_esw_sched_node *tc_arbiter_node, *vports_tc_node; + int err, num_tcs = esw_qos_num_tcs(vport_node->esw->dev); + u32 rate_limit_elem_ix; + + vport->qos.sched_nodes = kcalloc(num_tcs, + sizeof(struct mlx5_esw_sched_node *), + GFP_KERNEL); + if (!vport->qos.sched_nodes) { + NL_SET_ERR_MSG_MOD(extack, + "Allocating the vport TC scheduling elements failed."); + return -ENOMEM; + } + + rate_limit_elem_ix = type == SCHED_NODE_TYPE_RATE_LIMITER ? + vport_node->ix : 0; + tc_arbiter_node = type == SCHED_NODE_TYPE_RATE_LIMITER ? + vport_node->parent : vport_node; + list_for_each_entry(vports_tc_node, &tc_arbiter_node->children, entry) { + err = esw_qos_create_vport_tc_sched_node(vport, + rate_limit_elem_ix, + vports_tc_node, + extack); + if (err) + goto err_create_vport_tc; + } + + return 0; + +err_create_vport_tc: + esw_qos_destroy_vport_tc_sched_elements(vport, NULL); + + return err; +} + +static int +esw_qos_vport_tc_enable(struct mlx5_vport *vport, enum sched_node_type type, + struct netlink_ext_ack *extack) +{ + struct mlx5_esw_sched_node *vport_node = vport->qos.sched_node; + int err, new_level, max_level; + + if (type == SCHED_NODE_TYPE_TC_ARBITER_TSAR) { + /* Increase the parent's level by 2 to account for both the + * TC arbiter and the vports TC scheduling element. + */ + new_level = vport_node->parent->level + 2; + max_level = 1 << MLX5_CAP_QOS(vport_node->esw->dev, + log_esw_max_sched_depth); + if (new_level > max_level) { + NL_SET_ERR_MSG_MOD(extack, + "TC arbitration on leafs is not supported beyond max scheduling depth"); + return -EOPNOTSUPP; + } + } + + esw_assert_qos_lock_held(vport->dev->priv.eswitch); + + if (type == SCHED_NODE_TYPE_RATE_LIMITER) + err = esw_qos_create_rate_limit_element(vport_node, extack); + else + err = esw_qos_tc_arbiter_scheduling_setup(vport_node, extack); + if (err) + return err; + + /* Rate limiters impact multiple nodes not directly connected to them + * and are not direct members of the QoS hierarchy. + * Unlink it from the parent to reflect that. + */ + if (type == SCHED_NODE_TYPE_RATE_LIMITER) { + list_del_init(&vport_node->entry); + vport_node->level = 0; + } + + err = esw_qos_create_vport_tc_sched_elements(vport, type, extack); + if (err) + goto err_sched_nodes; + + return 0; + +err_sched_nodes: + if (type == SCHED_NODE_TYPE_RATE_LIMITER) { + esw_qos_node_destroy_sched_element(vport_node, NULL); + list_add_tail(&vport_node->entry, + &vport_node->parent->children); + vport_node->level = vport_node->parent->level + 1; + } else { + esw_qos_tc_arbiter_scheduling_teardown(vport_node, NULL); + } + return err; +} + +static void esw_qos_vport_tc_disable(struct mlx5_vport *vport, + struct netlink_ext_ack *extack) +{ + struct mlx5_esw_sched_node *vport_node = vport->qos.sched_node; + enum sched_node_type curr_type = vport_node->type; + + esw_qos_destroy_vport_tc_sched_elements(vport, extack); + + if (curr_type == SCHED_NODE_TYPE_RATE_LIMITER) + esw_qos_node_destroy_sched_element(vport_node, extack); + else + esw_qos_tc_arbiter_scheduling_teardown(vport_node, extack); +} + +static int esw_qos_set_vport_tcs_min_rate(struct mlx5_vport *vport, + u32 min_rate, + struct netlink_ext_ack *extack) +{ + struct mlx5_esw_sched_node *vport_node = vport->qos.sched_node; + int err, i, num_tcs = esw_qos_num_tcs(vport_node->esw->dev); + + for (i = 0; i < num_tcs; i++) { + err = esw_qos_set_node_min_rate(vport->qos.sched_nodes[i], + min_rate, extack); + if (err) + goto err_out; + } + vport_node->min_rate = min_rate; + + return 0; +err_out: + for (--i; i >= 0; i--) { + esw_qos_set_node_min_rate(vport->qos.sched_nodes[i], + vport_node->min_rate, extack); + } + return err; +} + static void esw_qos_vport_disable(struct mlx5_vport *vport, struct netlink_ext_ack *extack) { struct mlx5_esw_sched_node *vport_node = vport->qos.sched_node; struct mlx5_esw_sched_node *parent = vport_node->parent; + enum sched_node_type curr_type = vport_node->type;

- esw_qos_node_destroy_sched_element(vport_node, extack); + if (curr_type == SCHED_NODE_TYPE_VPORT) + esw_qos_node_destroy_sched_element(vport_node, extack); + else + esw_qos_vport_tc_disable(vport, extack);

vport_node->bw_share = 0; list_del_init(&vport_node->entry); @@ -631,7 +898,9 @@ static void esw_qos_vport_disable(struct mlx5_vport *vport, struct netlink_ext_a trace_mlx5_esw_vport_qos_destroy(vport_node->esw->dev, vport); }

-static int esw_qos_vport_enable(struct mlx5_vport *vport, struct mlx5_esw_sched_node *parent, +static int esw_qos_vport_enable(struct mlx5_vport *vport, + enum sched_node_type type, + struct mlx5_esw_sched_node *parent, struct netlink_ext_ack *extack) { int err; @@ -639,10 +908,16 @@ static int esw_qos_vport_enable(struct mlx5_vport *vport, struct mlx5_esw_sched_ esw_assert_qos_lock_held(vport->dev->priv.eswitch);

esw_qos_node_set_parent(vport->qos.sched_node, parent); - err = esw_qos_vport_create_sched_element(vport->qos.sched_node, extack); + if (type == SCHED_NODE_TYPE_VPORT) { + err = esw_qos_vport_create_sched_element(vport->qos.sched_node, + extack); + } else { + err = esw_qos_vport_tc_enable(vport, type, extack); + } if (err) return err;

+ vport->qos.sched_node->type = type; esw_qos_normalize_min_rate(parent->esw, parent, extack); trace_mlx5_esw_vport_qos_create(vport->dev, vport, vport->qos.sched_node->max_rate, @@ -673,9 +948,8 @@ static int mlx5_esw_qos_vport_enable(struct mlx5_vport *vport, enum sched_node_t sched_node->min_rate = min_rate; sched_node->vport = vport; vport->qos.sched_node = sched_node; - err = esw_qos_vport_enable(vport, parent, extack); + err = esw_qos_vport_enable(vport, type, parent, extack); if (err) { - __esw_qos_free_node(sched_node); esw_qos_put(esw); vport->qos.sched_node = NULL; } @@ -728,6 +1002,8 @@ static int mlx5_esw_qos_set_vport_min_rate(struct mlx5_vport *vport, u32 min_rat if (!vport_node) return mlx5_esw_qos_vport_enable(vport, SCHED_NODE_TYPE_VPORT, NULL, 0, min_rate, extack); + else if (vport_node->type == SCHED_NODE_TYPE_RATE_LIMITER) + return esw_qos_set_vport_tcs_min_rate(vport, min_rate, extack); else return esw_qos_set_node_min_rate(vport_node, min_rate, extack); } @@ -760,12 +1036,60 @@ bool mlx5_esw_qos_get_vport_rate(struct mlx5_vport *vport, u32 *max_rate, u32 *m return enabled; }

+static int esw_qos_vport_tc_check_type(enum sched_node_type curr_type, + enum sched_node_type new_type, + struct netlink_ext_ack *extack) +{ + if (curr_type == SCHED_NODE_TYPE_TC_ARBITER_TSAR && + new_type == SCHED_NODE_TYPE_RATE_LIMITER) { + NL_SET_ERR_MSG_MOD(extack, + "Cannot switch from vport-level TC arbitration to node-level TC arbitration"); + return -EOPNOTSUPP; + } + + if (curr_type == SCHED_NODE_TYPE_RATE_LIMITER && + new_type == SCHED_NODE_TYPE_TC_ARBITER_TSAR) { + NL_SET_ERR_MSG_MOD(extack, + "Cannot switch from node-level TC arbitration to vport-level TC arbitration"); + return -EOPNOTSUPP; + } + + return 0; +} + +static int esw_qos_vport_update(struct mlx5_vport *vport, + enum sched_node_type type, + struct mlx5_esw_sched_node *parent, + struct netlink_ext_ack *extack) +{ + struct mlx5_esw_sched_node *curr_parent = vport->qos.sched_node->parent; + enum sched_node_type curr_type = vport->qos.sched_node->type; + int err; + + esw_assert_qos_lock_held(vport->dev->priv.eswitch); + parent = parent ?: curr_parent; + if (curr_type == type && curr_parent == parent) + return 0; + + err = esw_qos_vport_tc_check_type(curr_type, type, extack); + if (err) + return err; + + esw_qos_vport_disable(vport, extack); + + err = esw_qos_vport_enable(vport, type, parent, extack); + if (err) + esw_qos_vport_enable(vport, curr_type, curr_parent, NULL); + + return err; +} + static int esw_qos_vport_update_parent(struct mlx5_vport *vport, struct mlx5_esw_sched_node *parent, struct netlink_ext_ack *extack) { struct mlx5_eswitch *esw = vport->dev->priv.eswitch; struct mlx5_esw_sched_node *curr_parent; - int err; + enum sched_node_type type;

esw_assert_qos_lock_held(esw); curr_parent = vport->qos.sched_node->parent; @@ -773,16 +1097,17 @@ static int esw_qos_vport_update_parent(struct mlx5_vport *vport, struct mlx5_esw if (curr_parent == parent) return 0;

- esw_qos_vport_disable(vport, extack); - - err = esw_qos_vport_enable(vport, parent, extack); - if (err) { - if (esw_qos_vport_enable(vport, curr_parent, NULL)) - esw_warn(parent->esw->dev, "vport restore QoS failed (vport=%d)\n", - vport->vport); - } + /* Set vport QoS type based on parent node type if different from + * default QoS; otherwise, use the vport's current QoS type. + */ + if (parent->type == SCHED_NODE_TYPE_TC_ARBITER_TSAR) + type = SCHED_NODE_TYPE_RATE_LIMITER; + else if (curr_parent->type == SCHED_NODE_TYPE_TC_ARBITER_TSAR) + type = SCHED_NODE_TYPE_VPORT; + else + type = vport->qos.sched_node->type;

- return err; + return esw_qos_vport_update(vport, type, parent, extack); }

static void @@ -1112,6 +1437,16 @@ static bool esw_qos_validate_unsupported_tc_bw(struct mlx5_eswitch *esw, return true; }

+static bool esw_qos_vport_validate_unsupported_tc_bw(struct mlx5_vport *vport, + u32 *tc_bw) +{ + struct mlx5_eswitch *esw = vport->qos.sched_node ? + vport->qos.sched_node->parent->esw : + vport->dev->priv.eswitch; + + return esw_qos_validate_unsupported_tc_bw(esw, tc_bw); +} + static bool esw_qos_tc_bw_disabled(u32 *tc_bw) { int i; @@ -1187,9 +1522,50 @@ int mlx5_esw_devlink_rate_leaf_tc_bw_set(struct devlink_rate *rate_leaf, u32 *tc_bw, struct netlink_ext_ack *extack) { - NL_SET_ERR_MSG_MOD(extack, - "TC bandwidth shares are not supported on leafs"); - return -EOPNOTSUPP; + struct mlx5_esw_sched_node *vport_node; + struct mlx5_vport *vport = priv; + struct mlx5_eswitch *esw; + bool disable; + int err = 0; + + esw = vport->dev->priv.eswitch; + if (!mlx5_esw_allowed(esw)) + return -EPERM; + + disable = esw_qos_tc_bw_disabled(tc_bw); + esw_qos_lock(esw); + + if (!esw_qos_vport_validate_unsupported_tc_bw(vport, tc_bw)) { + NL_SET_ERR_MSG_MOD(extack, + "E-Switch traffic classes number is not supported"); + err = -EOPNOTSUPP; + goto unlock; + } + + vport_node = vport->qos.sched_node; + if (disable && !vport_node) + goto unlock; + + if (disable) { + if (vport_node->type == SCHED_NODE_TYPE_TC_ARBITER_TSAR) + err = esw_qos_vport_update(vport, SCHED_NODE_TYPE_VPORT, + NULL, extack); + goto unlock; + } + + if (!vport_node) { + err = mlx5_esw_qos_vport_enable(vport, + SCHED_NODE_TYPE_TC_ARBITER_TSAR, + NULL, 0, 0, extack); + vport_node = vport->qos.sched_node; + } else { + err = esw_qos_vport_update(vport, + SCHED_NODE_TYPE_TC_ARBITER_TSAR, + NULL, extack); + } +unlock: + esw_qos_unlock(esw); + return err; }

int mlx5_esw_devlink_rate_node_tc_bw_set(struct devlink_rate *rate_node, @@ -1311,10 +1687,16 @@ int mlx5_esw_qos_vport_update_parent(struct mlx5_vport *vport, struct mlx5_esw_s }

esw_qos_lock(esw); - if (!vport->qos.sched_node && parent) - err = mlx5_esw_qos_vport_enable(vport, SCHED_NODE_TYPE_VPORT, parent, 0, 0, extack); - else if (vport->qos.sched_node) + if (!vport->qos.sched_node && parent) { + enum sched_node_type type; + + type = parent->type == SCHED_NODE_TYPE_TC_ARBITER_TSAR ? + SCHED_NODE_TYPE_RATE_LIMITER : SCHED_NODE_TYPE_VPORT; + err = mlx5_esw_qos_vport_enable(vport, type, parent, 0, 0, + extack); + } else if (vport->qos.sched_node) { err = esw_qos_vport_update_parent(vport, parent, extack); + } esw_qos_unlock(esw); return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 8573d36785f4..d59fdcb29cb8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -212,10 +212,20 @@ struct mlx5_vport {

struct mlx5_vport_info info;

- /* Protected with the E-Switch qos domain lock. */ + /* Protected with the E-Switch qos domain lock. The Vport QoS can + * either be disabled (sched_node is NULL) or in one of three states: + * 1. Regular QoS (sched_node is a vport node). + * 2. TC QoS enabled on the vport (sched_node is a TC arbiter). + * 3. TC QoS enabled on the vport's parent node + * (sched_node is a rate limit node). + * When TC is enabled in either mode, the vport owns vport TC scheduling + * nodes. + */ struct { - /* Vport scheduling element node. */ + /* Vport scheduling node. */ struct mlx5_esw_sched_node *sched_node; + /* Array of vport traffic class scheduling nodes. */ + struct mlx5_esw_sched_node **sched_nodes; } qos;

u16 vport;

From: Carolina Jubran cjubran@nvidia.com

This test suite validates the functionality of the devlink-rate API for traffic class (TC) bandwidth allocation. It ensures that bandwidth can be distributed between different traffic classes as configured, and verifies that explicit TC-to-queue mapping is required for the allocation to be effective.

The first test (test_no_tc_mapping_bandwidth) is marked as expected failure on mlx5, since the hardware automatically enforces traffic class separation by dynamically moving queues to the correct TC scheduler, even without explicit TC-to-queue mapping configuration.

Test output on mlx5: 1..2 # Created VF interface: eth5 # Created VLAN eth5.101 on eth5 with tc 3 and IP 198.51.100.2 # Created VLAN eth5.102 on eth5 with tc 4 and IP 198.51.100.10 # Set representor eth4 up and added to bridge # Bandwidth check results without TC mapping: # TC 3: 0.19 Gbits/sec # TC 4: 0.76 Gbits/sec # Total bandwidth: 0.95 Gbits/sec # TC 3 percentage: 20.0% # TC 4 percentage: 80.0% ok 1 devlink_rate_tc_bw.test_no_tc_mapping_bandwidth # XFAIL Bandwidth matched 80/20 split without TC mapping # Created VF interface: eth5 # Created VLAN eth5.101 on eth5 with tc 3 and IP 198.51.100.2 # Created VLAN eth5.102 on eth5 with tc 4 and IP 198.51.100.10 # Set representor eth4 up and added to bridge # Bandwidth check results with TC mapping: # TC 3: 0.21 Gbits/sec # TC 4: 0.78 Gbits/sec # Total bandwidth: 0.98 Gbits/sec # TC 3 percentage: 21.1% # TC 4 percentage: 78.9% # Bandwidth is distributed as 80/20 with TC mapping ok 2 devlink_rate_tc_bw.test_tc_mapping_bandwidth # Totals: pass:1 fail:0 xfail:1 xpass:0 skip:0 error:0

Signed-off-by: Carolina Jubran cjubran@nvidia.com Reviewed-by: Cosmin Ratiu cratiu@nvidia.com Reviewed-by: Nimrod Oren noren@nvidia.com Signed-off-by: Mark Bloch mbloch@nvidia.com --- Hi Jakub, you can find the iproute2 patch here: https://lore.kernel.org/netdev/20250625182545.86994-1-mbloch@nvidia.com/

.../drivers/net/hw/devlink_rate_tc_bw.py | 466 ++++++++++++++++++ .../testing/selftests/net/lib/py/__init__.py | 2 +- tools/testing/selftests/net/lib/py/ynl.py | 5 + 3 files changed, 472 insertions(+), 1 deletion(-) create mode 100755 tools/testing/selftests/drivers/net/hw/devlink_rate_tc_bw.py

diff --git a/tools/testing/selftests/drivers/net/hw/devlink_rate_tc_bw.py b/tools/testing/selftests/drivers/net/hw/devlink_rate_tc_bw.py new file mode 100755 index 000000000000..820d8a03becc --- /dev/null +++ b/tools/testing/selftests/drivers/net/hw/devlink_rate_tc_bw.py @@ -0,0 +1,466 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 + +""" +Devlink Rate TC Bandwidth Test Suite +=================================== + +This test suite verifies the functionality of devlink-rate traffic class (TC) +bandwidth distribution in a virtualized environment. The tests validate that +bandwidth can be properly allocated between different traffic classes and +that TC mapping works as expected. + +Test Environment: +---------------- +- Creates 1 VF +- Establishes a bridge connecting the VF representor and the uplink representor +- Sets up 2 VLAN interfaces on the VF with different VLAN IDs (101, 102) +- Configures different traffic classes (TC3 and TC4) for each VLAN + +Test Cases: +---------- +1. test_no_tc_mapping_bandwidth: + - Verifies that without TC mapping, bandwidth is NOT distributed according to + the configured 80/20 split between TC4 and TC3 + - This test should fail if bandwidth matches the 80/20 split without TC + mapping + - Expected: Bandwidth should NOT be distributed as 80/20 + +2. test_tc_mapping_bandwidth: + - Configures TC mapping using mqprio qdisc + - Verifies that with TC mapping, bandwidth IS distributed according to the + configured 80/20 split between TC3 and TC4 + - Expected: Bandwidth should be distributed as 80/20 + +Bandwidth Distribution: +---------------------- +- TC3 (VLAN 101): Configured for 80% of total bandwidth +- TC4 (VLAN 102): Configured for 20% of total bandwidth +- Total bandwidth: 1Gbps +- Tolerance: +-12% + +Hardware-Specific Behavior (mlx5): +-------------------------- +mlx5 hardware enforces traffic class separation by ensuring that each transmit +queue (SQ) is associated with a single TC. If a packet is sent on a queue that +doesn't match the expected TC (based on DSCP or VLAN priority and hypervisor-set +mapping), the hardware moves the queue to the correct TC scheduler to preserve +traffic isolation. + +This behavior means that even without explicit TC-to-queue mapping, bandwidth +enforcement may still appear to work���because the hardware dynamically adjusts +the scheduling context. However, this can lead to performance issues in high +rates and HOL blocking if traffic from different TCs is mixed on the same queue. +""" + +import json +import os +import subprocess +import threading +import time + +from lib.py import ksft_pr, ksft_run, ksft_exit +from lib.py import KsftSkipEx, KsftFailEx, KsftXfailEx +from lib.py import NetDrvEpEnv, DevlinkFamily +from lib.py import NlError +from lib.py import cmd, defer, ethtool, ip + + +class BandwidthValidator: + """ + Validates bandwidth totals and per-TC shares against expected values + with a tolerance. + """ + + def __init__(self): + self.tolerance_percent = 12 + self.expected_total_gbps = 1.0 + self.total_min_expected = self.min_expected(self.expected_total_gbps) + self.total_max_expected = self.max_expected(self.expected_total_gbps) + self.tc_expected_percent = { + 3: 20.0, + 4: 80.0, + } + + def min_expected(self, value): + """Calculates the minimum acceptable value based on tolerance.""" + return value - (value * self.tolerance_percent / 100) + + def max_expected(self, value): + """Calculates the maximum acceptable value based on tolerance.""" + return value + (value * self.tolerance_percent / 100) + + def bound(self, expected, value): + """Returns True if value is within expected tolerance.""" + return self.min_expected(expected) <= value <= self.max_expected(expected) + + def tc_bandwidth_bound(self, value, tc_ix): + """ + Returns True if the given bandwidth value is within tolerance + for the TC's expected bandwidth. + """ + expected = self.tc_expected_percent[tc_ix] + return self.bound(expected, value) + + +def setup_vf(cfg, set_tc_mapping=True): + """ + Sets up a VF on the given network interface. + + Enables SR-IOV and switchdev mode, brings the VF interface up, + and optionally configures TC mapping using mqprio. + """ + try: + cmd(f"devlink dev eswitch set pci/{cfg.pci} mode switchdev") + defer(cmd, f"devlink dev eswitch set pci/{cfg.pci} mode legacy") + except Exception as exc: + raise KsftSkipEx(f"Failed to enable switchdev mode on {cfg.pci}") from exc + try: + cmd(f"echo 1 > /sys/class/net/{cfg.ifname}/device/sriov_numvfs") + defer(cmd, f"echo 0 > /sys/class/net/{cfg.ifname}/device/sriov_numvfs") + except Exception as exc: + raise KsftSkipEx(f"Failed to enable SR-IOV on {cfg.ifname}") from exc + + time.sleep(2) + vf_ifc = (os.listdir( + f"/sys/class/net/{cfg.ifname}/device/virtfn0/net") or [None])[0] + if vf_ifc: + ip(f"link set dev {vf_ifc} up") + else: + raise KsftSkipEx("VF interface not found") + if set_tc_mapping: + cmd(f"tc qdisc add dev {vf_ifc} root handle 5 mqprio mode dcb hw 1 num_tc 8") + + return vf_ifc + + +def setup_vlans_on_vf(vf_ifc): + """ + Sets up two VLAN interfaces on the given VF, each mapped to a different TC. + """ + vlan_configs = [ + {"vlan_id": 101, "tc": 3, "ip": "198.51.100.2"}, + {"vlan_id": 102, "tc": 4, "ip": "198.51.100.10"}, + ] + + for config in vlan_configs: + vlan_dev = f"{vf_ifc}.{config['vlan_id']}" + ip(f"link add link {vf_ifc} name {vlan_dev} type vlan id {config['vlan_id']}") + ip(f"addr add {config['ip']}/29 dev {vlan_dev}") + ip(f"link set dev {vlan_dev} up") + ip(f"link set dev {vlan_dev} type vlan egress-qos-map 0:{config['tc']}") + ksft_pr(f"Created VLAN {vlan_dev} on {vf_ifc} with tc {config['tc']} and IP {config['ip']}") + + +def get_vf_info(cfg): + """ + Finds the VF representor interface and devlink port index + for the given PCI device used in the test environment. + """ + cfg.vf_representor = None + cfg.vf_port_index = None + out = subprocess.check_output(["devlink", "-j", "port", "show"], encoding="utf-8") + ports = json.loads(out)["port"] + + for port_name, props in ports.items(): + netdev = props.get("netdev") + + if (port_name.startswith(f"pci/{cfg.pci}/") and + props.get("vfnum") == 0): + cfg.vf_representor = netdev + cfg.vf_port_index = int(port_name.split("/")[-1]) + break + + +def setup_bridge(cfg): + """ + Creates and configures a Linux bridge, with both the uplink + and VF representor interfaces attached to it. + """ + bridge_name = f"br_{os.getpid()}" + ip(f"link add name {bridge_name} type bridge") + defer(cmd, f"ip link del name {bridge_name} type bridge") + + ip(f"link set dev {cfg.ifname} master {bridge_name}") + + rep_name = cfg.vf_representor + if rep_name: + ip(f"link set dev {rep_name} master {bridge_name}") + ip(f"link set dev {rep_name} up") + ksft_pr(f"Set representor {rep_name} up and added to bridge") + else: + raise KsftSkipEx("Could not find representor for the VF") + + ip(f"link set dev {bridge_name} up") + + +def setup_devlink_rate(cfg): + """ + Configures devlink rate tx_max and traffic class bandwidth for the VF. + """ + port_index = cfg.vf_port_index + if port_index is None: + raise KsftSkipEx("Could not find VF port index") + try: + cfg.devnl.rate_set({ + "bus-name": "pci", + "dev-name": cfg.pci, + "port-index": port_index, + "rate-tx-max": 125000000, + "rate-tc-bws": [ + {"rate-tc-index": 0, "rate-tc-bw": 0}, + {"rate-tc-index": 1, "rate-tc-bw": 0}, + {"rate-tc-index": 2, "rate-tc-bw": 0}, + {"rate-tc-index": 3, "rate-tc-bw": 20}, + {"rate-tc-index": 4, "rate-tc-bw": 80}, + {"rate-tc-index": 5, "rate-tc-bw": 0}, + {"rate-tc-index": 6, "rate-tc-bw": 0}, + {"rate-tc-index": 7, "rate-tc-bw": 0}, + ] + }) + except NlError as exc: + if exc.error == 95: # EOPNOTSUPP + raise KsftSkipEx("devlink rate configuration is not supported on the VF") from exc + raise KsftFailEx(f"rate_set failed on VF port {port_index}") from exc + + +def setup_remote_server(cfg): + """ + Sets up VLAN interfaces and starts iperf3 servers on the remote side. + """ + remote_dev = cfg.remote_ifname + vlan_ids = [101, 102] + remote_ips = ["198.51.100.1", "198.51.100.9"] + + for vlan_id, ip_addr in zip(vlan_ids, remote_ips): + vlan_dev = f"{remote_dev}.{vlan_id}" + cmd(f"ip link add link {remote_dev} name {vlan_dev} " + f"type vlan id {vlan_id}", host=cfg.remote) + cmd(f"ip addr add {ip_addr}/29 dev {vlan_dev}", host=cfg.remote) + cmd(f"ip link set dev {vlan_dev} up", host=cfg.remote) + cmd(f"iperf3 -s -1 -B {ip_addr}",background=True, host=cfg.remote) + defer(cmd, f"ip link del {vlan_dev}", host=cfg.remote) + + +def setup_test_environment(cfg, set_tc_mapping=True): + """ + Sets up the complete test environment including VF creation, VLANs, + bridge configuration, devlink rate setup, and the remote server. + """ + vf_ifc = setup_vf(cfg, set_tc_mapping) + ksft_pr(f"Created VF interface: {vf_ifc}") + + setup_vlans_on_vf(vf_ifc) + + get_vf_info(cfg) + setup_bridge(cfg) + + setup_devlink_rate(cfg) + setup_remote_server(cfg) + time.sleep(2) + + +def run_iperf_client(server_ip, local_ip, barrier, min_expected_gbps=0.1): + """ + Runs a single iperf3 client instance, binding to the given local IP. + Waits on a barrier to synchronize with other threads. + """ + try: + barrier.wait(timeout=10) + except Exception as exc: + raise KsftFailEx("iperf3 barrier wait timed") from exc + + iperf_cmd = ["iperf3", "-c", server_ip, "-B", local_ip, "-J"] + result = subprocess.run(iperf_cmd, capture_output=True, text=True, + check=True) + + try: + output = json.loads(result.stdout) + bits_per_second = output["end"]["sum_received"]["bits_per_second"] + gbps = bits_per_second / 1e9 + if gbps < min_expected_gbps: + ksft_pr( + f"iperf3 bandwidth too low: {gbps:.2f} Gbps " + f"(expected ��� {min_expected_gbps} Gbps)" + ) + return None + return gbps + except json.JSONDecodeError as exc: + ksft_pr(f"Failed to parse iperf3 JSON output: {exc}") + return None + + +def run_bandwidth_test(): + """ + Launches iperf3 client threads for each VLAN/TC pair and collects results. + """ + def _run_iperf_client_thread(server_ip, local_ip, results, barrier, tc_ix): + results[tc_ix] = run_iperf_client(server_ip, local_ip, barrier) + + vf_vlan_data = [ + # (local_ip, remote_ip, TC) + ("198.51.100.2", "198.51.100.1", 3), + ("198.51.100.10", "198.51.100.9", 4), + ] + + results = {} + threads = [] + start_barrier = threading.Barrier(len(vf_vlan_data)) + + for local_ip, remote_ip, tc_ix in vf_vlan_data: + thread = threading.Thread( + target=_run_iperf_client_thread, + args=(remote_ip, local_ip, results, start_barrier, tc_ix) + ) + thread.start() + threads.append(thread) + + for thread in threads: + thread.join() + + for tc_ix, tc_bw in results.items(): + if tc_bw is None: + raise KsftFailEx("iperf3 client failed; cannot evaluate bandwidth") + + return results + +def calculate_bandwidth_percentages(results): + """ + Calculates the percentage of total bandwidth received by TC3 and TC4. + """ + if 3 not in results or 4 not in results: + raise KsftFailEx(f"Missing expected TC results in {results}") + + tc3_bw = results[3] + tc4_bw = results[4] + total_bw = tc3_bw + tc4_bw + tc3_percentage = (tc3_bw / total_bw) * 100 + tc4_percentage = (tc4_bw / total_bw) * 100 + + return { + 'tc3_bw': tc3_bw, + 'tc4_bw': tc4_bw, + 'tc3_percentage': tc3_percentage, + 'tc4_percentage': tc4_percentage, + 'total_bw': total_bw + } + + +def print_bandwidth_results(bw_data, test_name): + """ + Prints bandwidth measurements and TC usage summary for a given test. + """ + ksft_pr(f"Bandwidth check results {test_name}:") + ksft_pr(f"TC 3: {bw_data['tc3_bw']:.2f} Gbits/sec") + ksft_pr(f"TC 4: {bw_data['tc4_bw']:.2f} Gbits/sec") + ksft_pr(f"Total bandwidth: {bw_data['total_bw']:.2f} Gbits/sec") + ksft_pr(f"TC 3 percentage: {bw_data['tc3_percentage']:.1f}%") + ksft_pr(f"TC 4 percentage: {bw_data['tc4_percentage']:.1f}%") + + +def verify_total_bandwidth(bw_data, validator): + """ + Ensures the total measured bandwidth falls within the acceptable tolerance. + """ + total = bw_data['total_bw'] + + if validator.bound(validator.expected_total_gbps, total): + return + + if total < validator.total_min_expected: + raise KsftSkipEx( + f"Total bandwidth {total:.2f} Gbps < minimum " + f"{validator.total_min_expected:.2f} Gbps; " + f"parent tx_max ({validator.expected_total_gbps:.1f} G) " + f"not reached, cannot validate share" + ) + + raise KsftFailEx( + f"Total bandwidth {total:.2f} Gbps exceeds allowed ceiling " + f"{validator.total_max_expected:.2f} Gbps " + f"(VF tx_max set to {validator.expected_total_gbps:.1f} G)" + ) + + +def check_bandwidth_distribution(bw_data, validator): + """ + Checks whether the measured TC3 and TC4 bandwidth percentages + fall within their expected tolerance ranges. + + Returns: + bool: True if both TC3 and TC4 percentages are within bounds. + """ + tc3_valid = validator.tc_bandwidth_bound(bw_data['tc3_percentage'], 3) + tc4_valid = validator.tc_bandwidth_bound(bw_data['tc4_percentage'], 4) + + return tc3_valid and tc4_valid + + +def run_bandwidth_distribution_test(cfg, set_tc_mapping): + """ + Runs parallel iperf3 tests for both TCs and collects results. + """ + setup_test_environment(cfg, set_tc_mapping) + bandwidths = run_bandwidth_test() + bw_data = calculate_bandwidth_percentages(bandwidths) + test_name = "with TC mapping" if set_tc_mapping else "without TC mapping" + print_bandwidth_results(bw_data, test_name) + + verify_total_bandwidth(bw_data, cfg.bw_validator) + + return check_bandwidth_distribution(bw_data, cfg.bw_validator) + + +def test_no_tc_mapping_bandwidth(cfg): + """ + Verifies that bandwidth is not split 80/20 without traffic class mapping. + """ + pass_bw_msg = "Bandwidth is NOT distributed as 80/20 without TC mapping" + fail_bw_msg = "Bandwidth matched 80/20 split without TC mapping" + is_mlx5 = "driver: mlx5" in ethtool(f"-i {cfg.ifname}").stdout + + if run_bandwidth_distribution_test(cfg, set_tc_mapping=False): + if is_mlx5: + raise KsftXfailEx(fail_bw_msg) + raise KsftFailEx(fail_bw_msg) + if is_mlx5: + raise KsftFailEx("mlx5 behavior changed:" + pass_bw_msg) + ksft_pr(pass_bw_msg) + + +def test_tc_mapping_bandwidth(cfg): + """ + Verifies that bandwidth is correctly split 80/20 between TC3 and TC4 + when traffic class mapping is set. + """ + if run_bandwidth_distribution_test(cfg, set_tc_mapping=True): + ksft_pr("Bandwidth is distributed as 80/20 with TC mapping") + else: + raise KsftFailEx("Bandwidth did not match 80/20 split with TC mapping") + + +def main() -> None: + """ + Main entry point for running the test cases. + """ + with NetDrvEpEnv(__file__, nsim_test=False) as cfg: + cfg.devnl = DevlinkFamily() + + cfg.pci = os.path.basename( + os.path.realpath(f"/sys/class/net/{cfg.ifname}/device") + ) + if not cfg.pci: + raise KsftSkipEx("Could not get PCI address of the interface") + cfg.require_cmd("iperf3") + cfg.require_cmd("iperf3", remote=True) + + cfg.bw_validator = BandwidthValidator() + + cases = [test_no_tc_mapping_bandwidth, test_tc_mapping_bandwidth] + + ksft_run(cases=cases, args=(cfg,)) + ksft_exit() + + +if __name__ == "__main__": + main() diff --git a/tools/testing/selftests/net/lib/py/__init__.py b/tools/testing/selftests/net/lib/py/__init__.py index 8697bd27dc30..02be28dcc089 100644 --- a/tools/testing/selftests/net/lib/py/__init__.py +++ b/tools/testing/selftests/net/lib/py/__init__.py @@ -6,4 +6,4 @@ from .netns import NetNS, NetNSEnter from .nsim import * from .utils import * from .ynl import NlError, YnlFamily, EthtoolFamily, NetdevFamily, RtnlFamily, RtnlAddrFamily -from .ynl import NetshaperFamily +from .ynl import NetshaperFamily, DevlinkFamily diff --git a/tools/testing/selftests/net/lib/py/ynl.py b/tools/testing/selftests/net/lib/py/ynl.py index 6329ae805abf..2b3a61ea3bfa 100644 --- a/tools/testing/selftests/net/lib/py/ynl.py +++ b/tools/testing/selftests/net/lib/py/ynl.py @@ -56,3 +56,8 @@ class NetshaperFamily(YnlFamily): def __init__(self, recv_size=0): super().__init__((SPEC_PATH / Path('net_shaper.yaml')).as_posix(), schema='', recv_size=recv_size) + +class DevlinkFamily(YnlFamily): + def __init__(self, recv_size=0): + super().__init__((SPEC_PATH / Path('devlink.yaml')).as_posix(), + schema='', recv_size=recv_size)