[dpdk-dev] [PATCH v4 2/2] ethdev: add traffic management API
Hemant Agrawal
hemant.agrawal at nxp.com
Wed May 24 13:28:17 CEST 2017
On 5/19/2017 10:42 PM, Cristian Dumitrescu wrote:
> This patch introduces the generic ethdev API for the traffic manager
> capability, which includes: hierarchical scheduling, traffic shaping,
> congestion management, packet marking.
>
> Main features:
> - Exposed as ethdev plugin capability (similar to rte_flow)
> - Capability query API per port, per level and per node
> - Scheduling algorithms: Strict Priority (SP), Weighed Fair Queuing (WFQ)
> - Traffic shaping: single/dual rate, private (per node) and shared (by
> multiple nodes) shapers
> - Congestion management for hierarchy leaf nodes: algorithms of tail drop,
> head drop, WRED; private (per node) and shared (by multiple nodes) WRED
> contexts
> - Packet marking: IEEE 802.1q (VLAN DEI), IETF RFC 3168 (IPv4/IPv6 ECN for
> TCP and SCTP), IETF RFC 2597 (IPv4 / IPv6 DSCP)
>
> Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu at intel.com>
> ---
> Changes in v4:
> - Implemented feedback from Hemant [6]
> - Capability API: Reworked the port, level and node capability API
> data structure to remove confusion due to "summary across all
> nodes" approach, which made it unclear whether a particular
> capability is supported by all nodes or by at least one node.
> - Capability API: Added flags for "all nodes have identical
> capability set"
> - Suspended state: documented the required behavior in Doxygen
> description
> - Implemented feedback from Jerin [7]
> - Node add: added level parameter (see new API function:
> rte_tm_node_add_check_level())
> - RTE_TM_ETH_FRAMING_OVERHEAD, RTE_TM_ETH_FRAMING_OVERHEAD_FCS:
> documented their usage in their Doxygen description
> - Capability API: for each function, mention the related
> capability field (Doxygen @see)
> - stats_mask, capability_mask: document the enum flags used to
> build each mask (Doxygen @see)
> - Rename rte_tm_get_leaf_nodes() to
> rte_tm_get_number_of_leaf_nodes()
> - Doxygen: add @param[in, out] to the description of all API funcs
> - Doxygen: fix hooks in doc/api/doxy-api-index.md
> - Rename rte_tm_hierarchy_set() to rte_tm_hierarchy_commit(), improved
> Doxygen description
> - Node add, node delete: improved Doxygen description
> - Fixed incorrect design assumption that packet-based weight mode for WFQ
> is identical to WRR. As result, removed all references to WRR support.
> Renamed the "scheduling mode" node parameters to "wfq_weight_mode".
>
> Changes in v3:
> - Implemented feedback from Jerin [5]
> - Changed naming convention: scheddev -> tm
> - Improvements on the capability API:
> - Specification of marking capabilities per color
> - WFQ/WRR groups: sp_n_children_max ->
> wfq_wrr_n_children_per_group_max, added wfq_wrr_n_groups_max,
> improved description of both, improved description of
> wfq_wrr_weight_max
> - Dynamic updates: added KEEP_LEVEL and CHANGE_LEVEL for parent
> update
> - Enforced/documented restrictions for root node (node_add() and
> update())
> - Enforced/documented shaper profile restrictions on PIR: PIR != 0,
> PIR >= CIR
> - Turned repetitive code in rte_tm.c into macro
> - Removed dependency on rte_red.h file (added RED params to rte_tm.h)
> - Color: removed "e_" from color names enum
> - Fixed small Doxygen style issues
>
> Changes in v2:
> - Implemented feedback from Hemant [4]
> - Improvements on the capability API
> - Added capability API for hierarchy level
> - Merged stats capability into the capability API
> - Added dynamic updates
> - Added non-leaf/leaf union to the node capability structure
> - Renamed sp_priority_min to sp_n_priorities_max, added
> clarifications
> - Fixed description for sp_n_children_max
> - Clarified and enforced rule on node ID range for leaf and non-leaf nodes
> - Added API functions to get node type (i.e. leaf/non-leaf):
> get_leaf_nodes(), node_type_get()
> - Added clarification for the root node: its creation, parent, role
> - Macro NODE_ID_NULL as root node's parent
> - Description of the node_add() and node_parent_update() API funcs
> - Added clarification for the first time add vs. subsequent updates rule
> - Cleaned up the description for the node_add() function
> - Statistics API improvements
> - Merged stats capability into the capability API
> - Added API function node_stats_update()
> - Added more stats per packet color
> - Added more error types
> - Fixed small Doxygen style issues
>
> Changes in v1 (since RFC [1]):
> - Implemented as ethdev plugin (similar to rte_flow) as opposed to more
> monolithic additions to ethdev itself
> - Implemented feedback from Jerin [2] and Hemant [3]. Implemented all the
> suggested items with only one exception, see the long list below,
> hopefully nothing was forgotten.
> - The item not done (hopefully for a good reason): driver-generated
> object IDs. IMO the choice to have application-generated object IDs
> adds marginal complexity to the driver (search ID function
> required), but it provides huge simplification for the application.
> The app does not need to worry about building & managing tree-like
> structure for storing driver-generated object IDs, the app can use
> its own convention for node IDs depending on the specific hierarchy
> that it needs. Trivial example: identify all level-2 nodes with IDs
> like 100, 200, 300, … and the level-3 nodes based on their level-2
> parents: 110, 120, 130, 140, …, 210, 220, 230, 240, …, 310, 320,
> 330, … and level-4 nodes based on their level-3 parents: 111, 112,
> 113, 114, …, 121, 122, 123, 124, …). Moreover, see the change log
> for the other related simplification that was implemented: leaf
> nodes now have predefined IDs that are the same with their Ethernet
> TX queue ID ( therefore no translation is required for leaf nodes).
> - Capability API. Done per port and per node as well.
> - Dual rate shapers
> - Added configuration of private shaper (per node) directly from the
> shaper profile as part of node API (no shaper ID needed for private
> shapers), while the shared shapers are configured outside of the node
> API using shaper profile and communicated to the node using shared
> shaper ID. So there is no configuration overhead for shared shapers if
> the app does not use any of them.
> - Leaf nodes now have predefined IDs that are the same with their Ethernet
> TX queue ID (therefore no translation is required for leaf nodes). This
> is also used to differentiate between a leaf node and a non-leaf node.
> - Domain-specific errors to give a precise indication of the error cause
> (same as done by rte_flow)
> - Packet marking API
> - Packet length optional adjustment for shapers, positive (e.g. for adding
> Ethernet framing overhead of 20 bytes) or negative (e.g. for rate
> limiting based on IP packet bytes)
>
> [1] RFC: http://dpdk.org/ml/archives/dev/2016-November/050956.html
> [2] Jerin’s feedback on RFC: http://www.dpdk.org/ml/archives/dev/2017-January/054484.html
> [3] Hemant’s feedback on RFC: http://www.dpdk.org/ml/archives/dev/2017-January/054866.html
> [4] Hemant's feedback on v1: http://www.dpdk.org/ml/archives/dev/2017-February/058033.html
> [5] Jerin's feedback on v1: http://www.dpdk.org/ml/archives/dev/2017-March/058895.html
> [6] Hemant's feedback on v3: http://www.dpdk.org/ml/archives/dev/2017-March/062354.html
> [7] Jerin's feedback on v3: http://www.dpdk.org/ml/archives/dev/2017-April/063429.html
>
> MAINTAINERS | 4 +
> lib/librte_ether/Makefile | 5 +-
> lib/librte_ether/rte_ether_version.map | 30 +
> lib/librte_ether/rte_tm.c | 448 ++++++++
> lib/librte_ether/rte_tm.h | 1923 ++++++++++++++++++++++++++++++++
> lib/librte_ether/rte_tm_driver.h | 373 +++++++
> 6 files changed, 2782 insertions(+), 1 deletion(-)
> create mode 100644 lib/librte_ether/rte_tm.c
> create mode 100644 lib/librte_ether/rte_tm.h
> create mode 100644 lib/librte_ether/rte_tm_driver.h
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index afb4cab..cdaf2ac 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -240,6 +240,10 @@ Flow API
> M: Adrien Mazarguil <adrien.mazarguil at 6wind.com>
> F: lib/librte_ether/rte_flow*
>
> +Traffic Management API
> +M: Cristian Dumitrescu <cristian.dumitrescu at intel.com>
> +F: lib/librte_ether/rte_tm*
> +
> Crypto API
> M: Declan Doherty <declan.doherty at intel.com>
> F: lib/librte_cryptodev/
> diff --git a/lib/librte_ether/Makefile b/lib/librte_ether/Makefile
> index 93fdde1..db692ae 100644
> --- a/lib/librte_ether/Makefile
> +++ b/lib/librte_ether/Makefile
> @@ -1,6 +1,6 @@
> # BSD LICENSE
> #
> -# Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
> +# Copyright(c) 2010-2017 Intel Corporation. All rights reserved.
> # All rights reserved.
> #
> # Redistribution and use in source and binary forms, with or without
> @@ -45,6 +45,7 @@ LIBABIVER := 6
>
> SRCS-y += rte_ethdev.c
> SRCS-y += rte_flow.c
> +SRCS-y += rte_tm.c
>
> #
> # Export include files
> @@ -56,5 +57,7 @@ SYMLINK-y-include += rte_eth_ctrl.h
> SYMLINK-y-include += rte_dev_info.h
> SYMLINK-y-include += rte_flow.h
> SYMLINK-y-include += rte_flow_driver.h
> +SYMLINK-y-include += rte_tm.h
> +SYMLINK-y-include += rte_tm_driver.h
>
> include $(RTE_SDK)/mk/rte.lib.mk
> diff --git a/lib/librte_ether/rte_ether_version.map b/lib/librte_ether/rte_ether_version.map
> index ff056e8..7f39904 100644
> --- a/lib/librte_ether/rte_ether_version.map
> +++ b/lib/librte_ether/rte_ether_version.map
> @@ -161,4 +161,34 @@ DPDK_17.08 {
> global:
>
> rte_eth_dev_tm_ops_get;
> + rte_tm_get_leaf_nodes;
> + rte_tm_node_type_get;
> + rte_tm_capabilities_get;
> + rte_tm_level_capabilities_get;
> + rte_tm_node_capabilities_get;
> + rte_tm_wred_profile_add;
> + rte_tm_wred_profile_delete;
> + rte_tm_shared_wred_context_add_update;
> + rte_tm_shared_wred_context_delete;
> + rte_tm_shaper_profile_add;
> + rte_tm_shaper_profile_delete;
> + rte_tm_shared_shaper_add_update;
> + rte_tm_shared_shaper_delete;
> + rte_tm_node_add;
> + rte_tm_node_delete;
> + rte_tm_node_suspend;
> + rte_tm_node_resume;
> + rte_tm_hierarchy_commit;
> + rte_tm_node_parent_update;
> + rte_tm_node_shaper_update;
> + rte_tm_node_shared_shaper_update;
> + rte_tm_node_stats_update;
> + rte_tm_node_wfq_weight_mode_update;
> + rte_tm_node_cman_update;
> + rte_tm_node_wred_context_update;
> + rte_tm_node_shared_wred_context_update;
> + rte_tm_node_stats_read;
> + rte_tm_mark_vlan_dei;
> + rte_tm_mark_ip_ecn;
> + rte_tm_mark_ip_dscp;
> } DPDK_17.05
> diff --git a/lib/librte_ether/rte_tm.c b/lib/librte_ether/rte_tm.c
> new file mode 100644
> index 0000000..2617a1a
> --- /dev/null
> +++ b/lib/librte_ether/rte_tm.c
> @@ -0,0 +1,448 @@
> +/*-
> + * BSD LICENSE
> + *
> + * Copyright(c) 2017 Intel Corporation. All rights reserved.
> + * All rights reserved.
> + *
> + * Redistribution and use in source and binary forms, with or without
> + * modification, are permitted provided that the following conditions
> + * are met:
> + *
> + * * Redistributions of source code must retain the above copyright
> + * notice, this list of conditions and the following disclaimer.
> + * * Redistributions in binary form must reproduce the above copyright
> + * notice, this list of conditions and the following disclaimer in
> + * the documentation and/or other materials provided with the
> + * distribution.
> + * * Neither the name of Intel Corporation nor the names of its
> + * contributors may be used to endorse or promote products derived
> + * from this software without specific prior written permission.
> + *
> + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
> + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
> + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
> + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
> + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
> + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
> + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
> + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + */
> +
> +#include <stdint.h>
> +
> +#include <rte_errno.h>
> +#include "rte_ethdev.h"
> +#include "rte_tm_driver.h"
> +#include "rte_tm.h"
> +
> +/* Get generic traffic manager operations structure from a port. */
> +const struct rte_tm_ops *
> +rte_tm_ops_get(uint8_t port_id, struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + const struct rte_tm_ops *ops;
> +
> + if (!rte_eth_dev_is_valid_port(port_id)) {
> + rte_tm_error_set(error,
> + ENODEV,
> + RTE_TM_ERROR_TYPE_UNSPECIFIED,
> + NULL,
> + rte_strerror(ENODEV));
> + return NULL;
> + }
> +
> + if ((dev->dev_ops->tm_ops_get == NULL) ||
> + (dev->dev_ops->tm_ops_get(dev, &ops) != 0) ||
> + (ops == NULL)) {
> + rte_tm_error_set(error,
> + ENOSYS,
> + RTE_TM_ERROR_TYPE_UNSPECIFIED,
> + NULL,
> + rte_strerror(ENOSYS));
> + return NULL;
> + }
> +
> + return ops;
> +}
> +
> +#define RTE_TM_FUNC(port_id, func) \
> +({ \
> + const struct rte_tm_ops *ops = \
> + rte_tm_ops_get(port_id, error); \
> + if (ops == NULL) \
> + return -rte_errno; \
> + \
> + if (ops->func == NULL) \
> + return -rte_tm_error_set(error, \
> + ENOSYS, \
> + RTE_TM_ERROR_TYPE_UNSPECIFIED, \
> + NULL, \
> + rte_strerror(ENOSYS)); \
> + \
> + ops->func; \
> +})
> +
> +/* Get number of leaf nodes */
> +int
> +rte_tm_get_number_of_leaf_nodes(uint8_t port_id,
> + uint32_t *n_leaf_nodes,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + const struct rte_tm_ops *ops =
> + rte_tm_ops_get(port_id, error);
> +
> + if (ops == NULL)
> + return -rte_errno;
> +
> + if (n_leaf_nodes == NULL) {
> + rte_tm_error_set(error,
> + EINVAL,
> + RTE_TM_ERROR_TYPE_UNSPECIFIED,
> + NULL,
> + rte_strerror(EINVAL));
> + return -rte_errno;
> + }
> +
> + *n_leaf_nodes = dev->data->nb_tx_queues;
> + return 0;
> +}
> +
> +/* Check node type (leaf or non-leaf) */
> +int
> +rte_tm_node_type_get(uint8_t port_id,
> + uint32_t node_id,
> + int *is_leaf,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_type_get)(dev,
> + node_id, is_leaf, error);
> +}
> +
> +/* Get node level */
> +int
> +rte_tm_node_level_get(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t *level_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_level_get)(dev,
> + node_id, level_id, error);
> +}
> +
> +/* Get capabilities */
> +int rte_tm_capabilities_get(uint8_t port_id,
> + struct rte_tm_capabilities *cap,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, capabilities_get)(dev,
> + cap, error);
> +}
> +
> +/* Get level capabilities */
> +int rte_tm_level_capabilities_get(uint8_t port_id,
> + uint32_t level_id,
> + struct rte_tm_level_capabilities *cap,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, level_capabilities_get)(dev,
> + level_id, cap, error);
> +}
> +
> +/* Get node capabilities */
> +int rte_tm_node_capabilities_get(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_node_capabilities *cap,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_capabilities_get)(dev,
> + node_id, cap, error);
> +}
> +
> +/* Add WRED profile */
> +int rte_tm_wred_profile_add(uint8_t port_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_wred_params *profile,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, wred_profile_add)(dev,
> + wred_profile_id, profile, error);
> +}
> +
> +/* Delete WRED profile */
> +int rte_tm_wred_profile_delete(uint8_t port_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, wred_profile_delete)(dev,
> + wred_profile_id, error);
> +}
> +
> +/* Add/update shared WRED context */
> +int rte_tm_shared_wred_context_add_update(uint8_t port_id,
> + uint32_t shared_wred_context_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, shared_wred_context_add_update)(dev,
> + shared_wred_context_id, wred_profile_id, error);
> +}
> +
> +/* Delete shared WRED context */
> +int rte_tm_shared_wred_context_delete(uint8_t port_id,
> + uint32_t shared_wred_context_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, shared_wred_context_delete)(dev,
> + shared_wred_context_id, error);
> +}
> +
> +/* Add shaper profile */
> +int rte_tm_shaper_profile_add(uint8_t port_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_shaper_params *profile,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, shaper_profile_add)(dev,
> + shaper_profile_id, profile, error);
> +}
> +
> +/* Delete WRED profile */
> +int rte_tm_shaper_profile_delete(uint8_t port_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, shaper_profile_delete)(dev,
> + shaper_profile_id, error);
> +}
> +
> +/* Add shared shaper */
> +int rte_tm_shared_shaper_add_update(uint8_t port_id,
> + uint32_t shared_shaper_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, shared_shaper_add_update)(dev,
> + shared_shaper_id, shaper_profile_id, error);
> +}
> +
> +/* Delete shared shaper */
> +int rte_tm_shared_shaper_delete(uint8_t port_id,
> + uint32_t shared_shaper_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, shared_shaper_delete)(dev,
> + shared_shaper_id, error);
> +}
> +
> +/* Add node to port traffic manager hierarchy */
> +int rte_tm_node_add(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t parent_node_id,
> + uint32_t priority,
> + uint32_t weight,
> + struct rte_tm_node_params *params,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_add)(dev,
> + node_id, parent_node_id, priority, weight, params, error);
> +}
> +
> +/* Delete node from traffic manager hierarchy */
> +int rte_tm_node_delete(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_delete)(dev,
> + node_id, error);
> +}
> +
> +/* Suspend node */
> +int rte_tm_node_suspend(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_suspend)(dev,
> + node_id, error);
> +}
> +
> +/* Resume node */
> +int rte_tm_node_resume(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_resume)(dev,
> + node_id, error);
> +}
> +
> +/* Commit the initial port traffic manager hierarchy */
> +int rte_tm_hierarchy_commit(uint8_t port_id,
> + int clear_on_fail,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, hierarchy_commit)(dev,
> + clear_on_fail, error);
> +}
> +
> +/* Update node parent */
> +int rte_tm_node_parent_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t parent_node_id,
> + uint32_t priority,
> + uint32_t weight,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_parent_update)(dev,
> + node_id, parent_node_id, priority, weight, error);
> +}
> +
> +/* Update node private shaper */
> +int rte_tm_node_shaper_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_shaper_update)(dev,
> + node_id, shaper_profile_id, error);
> +}
> +
> +/* Update node shared shapers */
> +int rte_tm_node_shared_shaper_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t shared_shaper_id,
> + int add,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_shared_shaper_update)(dev,
> + node_id, shared_shaper_id, add, error);
> +}
> +
> +/* Update node stats */
> +int rte_tm_node_stats_update(uint8_t port_id,
> + uint32_t node_id,
> + uint64_t stats_mask,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_stats_update)(dev,
> + node_id, stats_mask, error);
> +}
> +
> +/* Update WFQ weight mode */
> +int rte_tm_node_wfq_weight_mode_update(uint8_t port_id,
> + uint32_t node_id,
> + int *wfq_weight_mode,
> + uint32_t n_sp_priorities,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_wfq_weight_mode_update)(dev,
> + node_id, wfq_weight_mode, n_sp_priorities, error);
> +}
> +
> +/* Update node congestion management mode */
> +int rte_tm_node_cman_update(uint8_t port_id,
> + uint32_t node_id,
> + enum rte_tm_cman_mode cman,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_cman_update)(dev,
> + node_id, cman, error);
> +}
> +
> +/* Update node private WRED context */
> +int rte_tm_node_wred_context_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_wred_context_update)(dev,
> + node_id, wred_profile_id, error);
> +}
> +
> +/* Update node shared WRED context */
> +int rte_tm_node_shared_wred_context_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t shared_wred_context_id,
> + int add,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_shared_wred_context_update)(dev,
> + node_id, shared_wred_context_id, add, error);
> +}
> +
> +/* Read and/or clear stats counters for specific node */
> +int rte_tm_node_stats_read(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_node_stats *stats,
> + uint64_t *stats_mask,
> + int clear,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, node_stats_read)(dev,
> + node_id, stats, stats_mask, clear, error);
> +}
> +
> +/* Packet marking - VLAN DEI */
> +int rte_tm_mark_vlan_dei(uint8_t port_id,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, mark_vlan_dei)(dev,
> + mark_green, mark_yellow, mark_red, error);
> +}
> +
> +/* Packet marking - IPv4/IPv6 ECN */
> +int rte_tm_mark_ip_ecn(uint8_t port_id,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, mark_ip_ecn)(dev,
> + mark_green, mark_yellow, mark_red, error);
> +}
> +
> +/* Packet marking - IPv4/IPv6 DSCP */
> +int rte_tm_mark_ip_dscp(uint8_t port_id,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error)
> +{
> + struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> + return RTE_TM_FUNC(port_id, mark_ip_dscp)(dev,
> + mark_green, mark_yellow, mark_red, error);
> +}
> diff --git a/lib/librte_ether/rte_tm.h b/lib/librte_ether/rte_tm.h
> new file mode 100644
> index 0000000..22167c2
> --- /dev/null
> +++ b/lib/librte_ether/rte_tm.h
> @@ -0,0 +1,1923 @@
> +/*-
> + * BSD LICENSE
> + *
> + * Copyright(c) 2017 Intel Corporation. All rights reserved.
> + * All rights reserved.
> + *
> + * Redistribution and use in source and binary forms, with or without
> + * modification, are permitted provided that the following conditions
> + * are met:
> + *
> + * * Redistributions of source code must retain the above copyright
> + * notice, this list of conditions and the following disclaimer.
> + * * Redistributions in binary form must reproduce the above copyright
> + * notice, this list of conditions and the following disclaimer in
> + * the documentation and/or other materials provided with the
> + * distribution.
> + * * Neither the name of Intel Corporation nor the names of its
> + * contributors may be used to endorse or promote products derived
> + * from this software without specific prior written permission.
> + *
> + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
> + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
> + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
> + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
> + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
> + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
> + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
> + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + */
> +
> +#ifndef __INCLUDE_RTE_TM_H__
> +#define __INCLUDE_RTE_TM_H__
> +
> +/**
> + * @file
> + * RTE Generic Traffic Manager API
> + *
> + * This interface provides the ability to configure the traffic manager in a
> + * generic way. It includes features such as: hierarchical scheduling,
> + * traffic shaping, congestion management, packet marking, etc.
> + */
> +
> +#include <stdint.h>
> +
> +#ifdef __cplusplus
> +extern "C" {
> +#endif
> +
> +/**
> + * Ethernet framing overhead.
> + *
> + * Overhead fields per Ethernet frame:
> + * 1. Preamble: 7 bytes;
> + * 2. Start of Frame Delimiter (SFD): 1 byte;
> + * 3. Inter-Frame Gap (IFG): 12 bytes.
> + *
> + * One of the typical values for the *pkt_length_adjust* field of the shaper
> + * profile.
> + *
> + * @see struct rte_tm_shaper_params
> + *
> + */
> +#define RTE_TM_ETH_FRAMING_OVERHEAD 20
> +
> +/**
> + * Ethernet framing overhead including the Frame Check Sequence (FCS) field.
> + * Useful when FCS is generated and added at the end of the Ethernet frame on
> + * TX side without any SW intervention.
> + *
> + * One of the typical values for the pkt_length_adjust field of the shaper
> + * profile.
> + *
> + * @see struct rte_tm_shaper_params
> + */
> +#define RTE_TM_ETH_FRAMING_OVERHEAD_FCS 24
> +
> +/**< Invalid WRED profile ID */
> +#define RTE_TM_WRED_PROFILE_ID_NONE UINT32_MAX
> +
> +/**< Invalid shaper profile ID */
> +#define RTE_TM_SHAPER_PROFILE_ID_NONE UINT32_MAX
> +
> +/**< Node ID for the parent of the root node */
> +#define RTE_TM_NODE_ID_NULL UINT32_MAX
> +
> +/**
> + * Color
> + */
> +enum rte_tm_color {
> + RTE_TM_GREEN = 0, /**< Green */
> + RTE_TM_YELLOW, /**< Yellow */
> + RTE_TM_RED, /**< Red */
> + RTE_TM_COLORS /**< Number of colors */
> +};
> +
> +/**
> + * Node statistics counter type
> + */
> +enum rte_tm_stats_type {
> + /**< Number of packets scheduled from current node. */
> + RTE_TM_STATS_N_PKTS = 1 << 0,
> +
> + /**< Number of bytes scheduled from current node. */
> + RTE_TM_STATS_N_BYTES = 1 << 1,
> +
> + /**< Number of green packets dropped by current leaf node. */
> + RTE_TM_STATS_N_PKTS_GREEN_DROPPED = 1 << 2,
> +
> + /**< Number of yellow packets dropped by current leaf node. */
> + RTE_TM_STATS_N_PKTS_YELLOW_DROPPED = 1 << 3,
> +
> + /**< Number of red packets dropped by current leaf node. */
> + RTE_TM_STATS_N_PKTS_RED_DROPPED = 1 << 4,
> +
> + /**< Number of green bytes dropped by current leaf node. */
> + RTE_TM_STATS_N_BYTES_GREEN_DROPPED = 1 << 5,
> +
> + /**< Number of yellow bytes dropped by current leaf node. */
> + RTE_TM_STATS_N_BYTES_YELLOW_DROPPED = 1 << 6,
> +
> + /**< Number of red bytes dropped by current leaf node. */
> + RTE_TM_STATS_N_BYTES_RED_DROPPED = 1 << 7,
> +
> + /**< Number of packets currently waiting in the packet queue of current
> + * leaf node.
> + */
> + RTE_TM_STATS_N_PKTS_QUEUED = 1 << 8,
> +
> + /**< Number of bytes currently waiting in the packet queue of current
> + * leaf node.
> + */
> + RTE_TM_STATS_N_BYTES_QUEUED = 1 << 9,
> +};
> +
> +/**
> + * Node statistics counters
> + */
> +struct rte_tm_node_stats {
> + /**< Number of packets scheduled from current node. */
> + uint64_t n_pkts;
> +
> + /**< Number of bytes scheduled from current node. */
> + uint64_t n_bytes;
> +
> + /**< Statistics counters for leaf nodes only. */
> + struct {
> + /**< Number of packets dropped by current leaf node per each
> + * color.
> + */
> + uint64_t n_pkts_dropped[RTE_TM_COLORS];
> +
> + /**< Number of bytes dropped by current leaf node per each
> + * color.
> + */
> + uint64_t n_bytes_dropped[RTE_TM_COLORS];
> +
> + /**< Number of packets currently waiting in the packet queue of
> + * current leaf node.
> + */
> + uint64_t n_pkts_queued;
> +
> + /**< Number of bytes currently waiting in the packet queue of
> + * current leaf node.
> + */
> + uint64_t n_bytes_queued;
> + } leaf;
> +};
> +
> +/**
> + * Traffic manager dynamic updates
> + */
> +enum rte_tm_dynamic_update_type {
> + /**< Dynamic parent node update. The new parent node is located on same
> + * hierarchy level as the former parent node. Consequently, the node
> + * whose parent is changed preserves its hierarchy level.
> + */
> + RTE_TM_UPDATE_NODE_PARENT_KEEP_LEVEL = 1 << 0,
> +
> + /**< Dynamic parent node update. The new parent node is located on
> + * different hierarchy level than the former parent node. Consequently,
> + * the node whose parent is changed also changes its hierarchy level.
> + */
> + RTE_TM_UPDATE_NODE_PARENT_CHANGE_LEVEL = 1 << 1,
> +
> + /**< Dynamic node add/delete. */
> + RTE_TM_UPDATE_NODE_ADD_DELETE = 1 << 2,
> +
> + /**< Suspend/resume nodes. */
> + RTE_TM_UPDATE_NODE_SUSPEND_RESUME = 1 << 3,
> +
> + /**< Dynamic switch between byte-based and packet-based WFQ weights. */
> + RTE_TM_UPDATE_NODE_WFQ_WEIGHT_MODE = 1 << 4,
> +
> + /**< Dynamic update on number of SP priorities. */
> + RTE_TM_UPDATE_NODE_N_SP_PRIORITIES = 1 << 5,
> +
> + /**< Dynamic update of congestion management mode for leaf nodes. */
> + RTE_TM_UPDATE_NODE_CMAN = 1 << 6,
> +
> + /**< Dynamic update of the set of enabled stats counter types. */
> + RTE_TM_UPDATE_NODE_STATS = 1 << 7,
> +};
> +
> +/**
> + * Traffic manager capabilities
> + */
> +struct rte_tm_capabilities {
> + /**< Maximum number of nodes. */
> + uint32_t n_nodes_max;
> +
> + /**< Maximum number of levels (i.e. number of nodes connecting the root
> + * node with any leaf node, including the root and the leaf).
> + */
> + uint32_t n_levels_max;
> +
> + /**< When non-zero, this flag indicates that all the non-leaf nodes
> + * (with the exception of the root node) have identical capability set.
> + */
> + int non_leaf_nodes_identical;
> +
> + /**< When non-zero, this flag indicates that all the leaf nodes have
> + * identical capability set.
> + */
> + int leaf_nodes_identical;
> +
> + /**< Maximum number of shapers, either private or shared. In case the
> + * implementation does not share any resources between private and
> + * shared shapers, it is typically equal to the sum of
> + * *shaper_private_n_max* and *shaper_shared_n_max*.
> + */
> + uint32_t shaper_n_max;
> +
> + /**< Maximum number of private shapers. Indicates the maximum number of
> + * nodes that can concurrently have their private shaper enabled.
> + */
> + uint32_t shaper_private_n_max;
> +
> + /**< Maximum number of private shapers that support dual rate shaping.
> + * Indicates the maximum number of nodes that can concurrently have
> + * their private shaper enabled with dual rate support. Only valid when
> + * private shapers are supported. The value of zero indicates that dual
> + * rate shaping is not available for private shapers. The maximum value
> + * is *shaper_private_n_max*.
> + */
> + int shaper_private_dual_rate_n_max;
> +
> + /**< Minimum committed/peak rate (bytes per second) for any private
> + * shaper. Valid only when private shapers are supported.
> + */
> + uint64_t shaper_private_rate_min;
> +
> + /**< Maximum committed/peak rate (bytes per second) for any private
> + * shaper. Valid only when private shapers are supported.
> + */
> + uint64_t shaper_private_rate_max;
> +
> + /**< Maximum number of shared shapers. The value of zero indicates that
> + * shared shapers are not supported.
> + */
> + uint32_t shaper_shared_n_max;
> +
> + /**< Maximum number of nodes that can share the same shared shaper.
> + * Only valid when shared shapers are supported.
> + */
> + uint32_t shaper_shared_n_nodes_per_shaper_max;
> +
> + /**< Maximum number of shared shapers a node can be part of. This
> + * parameter indicates that there is at least one node that can be
> + * configured with this many shared shapers, which might not be true for
> + * all the nodes. Only valid when shared shapers are supported, in which
> + * case it ranges from 1 to *shaper_shared_n_max*.
> + */
> + uint32_t shaper_shared_n_shapers_per_node_max;
> +
> + /**< Maximum number of shared shapers that can be configured with dual
> + * rate shaping. The value of zero indicates that dual rate shaping
> + * support is not available for shared shapers.
> + */
> + uint32_t shaper_shared_dual_rate_n_max;
> +
> + /**< Minimum committed/peak rate (bytes per second) for any shared
> + * shaper. Only valid when shared shapers are supported.
> + */
> + uint64_t shaper_shared_rate_min;
> +
> + /**< Maximum committed/peak rate (bytes per second) for any shared
> + * shaper. Only valid when shared shapers are supported.
> + */
> + uint64_t shaper_shared_rate_max;
> +
> + /**< Minimum value allowed for packet length adjustment for any private
> + * or shared shaper.
> + */
> + int shaper_pkt_length_adjust_min;
> +
> + /**< Maximum value allowed for packet length adjustment for any private
> + * or shared shaper.
> + */
> + int shaper_pkt_length_adjust_max;
> +
> + /**< Maximum number of children nodes. This parameter indicates that
> + * there is at least one non-leaf node that can be configured with this
> + * many children nodes, which might not be true for all the non-leaf
> + * nodes.
> + */
> + uint32_t sched_n_children_max;
> +
> + /**< Maximum number of supported priority levels. This parameter
> + * indicates that there is at least one non-leaf node that can be
> + * configured with this many priority levels for managing its children
> + * nodes, which might not be true for all the non-leaf nodes. The value
> + * of zero is invalid. The value of 1 indicates that only priority 0 is
> + * supported, which essentially means that Strict Priority (SP)
> + * algorithm is not supported.
> + */
> + uint32_t sched_sp_n_priorities_max;
> +
> + /**< Maximum number of sibling nodes that can have the same priority at
> + * any given time, i.e. maximum size of the WFQ sibling node group. This
> + * parameter indicates there is at least one non-leaf node that meets
> + * this condition, which might not be true for all the non-leaf nodes.
> + * The value of zero is invalid. The value of 1 indicates that WFQ
> + * algorithm is not supported. The maximum value is
> + * *sched_n_children_max*.
> + */
> + uint32_t sched_wfq_n_children_per_group_max;
> +
> + /**< Maximum number of priority levels that can have more than one child
> + * node at any given time, i.e. maximum number of WFQ sibling node
> + * groups that have two or more members. This parameter indicates there
> + * is at least one non-leaf node that meets this condition, which might
> + * not be true for all the non-leaf nodes. The value of zero states that
> + * WFQ algorithm is not supported. The value of 1 indicates that
> + * (*sched_sp_n_priorities_max* - 1) priority levels have at most one
> + * child node, so there can be only one priority level with two or
> + * more sibling nodes making up a WFQ group. The maximum value is:
> + * min(floor(*sched_n_children_max* / 2), *sched_sp_n_priorities_max*).
> + */
> + uint32_t sched_wfq_n_groups_max;
> +
> + /**< Maximum WFQ weight. The value of 1 indicates that all sibling nodes
> + * with same priority have the same WFQ weight, so WFQ is reduced to FQ.
> + */
> + uint32_t sched_wfq_weight_max;
> +
> + /**< Head drop algorithm support. When non-zero, this parameter
> + * indicates that there is at least one leaf node that supports the head
> + * drop algorithm, which might not be true for all the leaf nodes.
> + */
> + int cman_head_drop_supported;
> +
> + /**< Maximum number of WRED contexts, either private or shared. In case
> + * the implementation does not share any resources between private and
> + * shared WRED contexts, it is typically equal to the sum of
> + * *cman_wred_context_private_n_max* and
> + * *cman_wred_context_shared_n_max*.
> + */
> + uint32_t cman_wred_context_n_max;
> +
> + /**< Maximum number of private WRED contexts. Indicates the maximum
> + * number of leaf nodes that can concurrently have their private WRED
> + * context enabled.
> + */
> + uint32_t cman_wred_context_private_n_max;
> +
> + /**< Maximum number of shared WRED contexts. The value of zero
> + * indicates that shared WRED contexts are not supported.
> + */
> + uint32_t cman_wred_context_shared_n_max;
> +
> + /**< Maximum number of leaf nodes that can share the same WRED context.
> + * Only valid when shared WRED contexts are supported.
> + */
> + uint32_t cman_wred_context_shared_n_nodes_per_context_max;
> +
> + /**< Maximum number of shared WRED contexts a leaf node can be part of.
> + * This parameter indicates that there is at least one leaf node that
> + * can be configured with this many shared WRED contexts, which might
> + * not be true for all the leaf nodes. Only valid when shared WRED
> + * contexts are supported, in which case it ranges from 1 to
> + * *cman_wred_context_shared_n_max*.
> + */
> + uint32_t cman_wred_context_shared_n_contexts_per_node_max;
> +
> + /**< Support for VLAN DEI packet marking (per color). */
> + int mark_vlan_dei_supported[RTE_TM_COLORS];
> +
> + /**< Support for IPv4/IPv6 ECN marking of TCP packets (per color). */
> + int mark_ip_ecn_tcp_supported[RTE_TM_COLORS];
> +
> + /**< Support for IPv4/IPv6 ECN marking of SCTP packets (per color). */
> + int mark_ip_ecn_sctp_supported[RTE_TM_COLORS];
> +
> + /**< Support for IPv4/IPv6 DSCP packet marking (per color). */
> + int mark_ip_dscp_supported[RTE_TM_COLORS];
> +
> + /**< Set of supported dynamic update operations.
> + * @see enum rte_tm_dynamic_update_type
> + */
> + uint64_t dynamic_update_mask;
> +
> + /**< Set of supported statistics counter types.
> + * @see enum rte_tm_stats_type
> + */
> + uint64_t stats_mask;
> +};
> +
> +/**
> + * Traffic manager level capabilities
> + */
> +struct rte_tm_level_capabilities {
> + /**< Maximum number of nodes for the current hierarchy level. */
> + uint32_t n_nodes_max;
> +
> + /**< Maximum number of non-leaf nodes for the current hierarchy level.
> + * The value of 0 indicates that current level only supports leaf
> + * nodes. The maximum value is *n_nodes_max*.
> + */
> + uint32_t n_nodes_nonleaf_max;
> +
> + /**< Maximum number of leaf nodes for the current hierarchy level. The
> + * value of 0 indicates that current level only supports non-leaf
> + * nodes. The maximum value is *n_nodes_max*.
> + */
> + uint32_t n_nodes_leaf_max;
> +
> + /**< When non-zero, this flag indicates that all the non-leaf nodes on
> + * this level have identical capability set. Valid only when
> + * *n_nodes_nonleaf_max* is non-zero.
> + */
> + int non_leaf_nodes_identical;
> +
> + /**< When non-zero, this flag indicates that all the leaf nodes on this
> + * level have identical capability set. Valid only when
> + * *n_nodes_leaf_max* is non-zero.
> + */
> + int leaf_nodes_identical;
> +
> + union {
> + /**< Items valid only for the non-leaf nodes on this level. */
> + struct {
> + /**< Private shaper support. When non-zero, it indicates
> + * there is at least one non-leaf node on this level
> + * with private shaper support, which may not be the
> + * case for all the non-leaf nodes on this level.
> + */
> + int shaper_private_supported;
> +
> + /**< Dual rate support for private shaper. Valid only
> + * when private shaper is supported for the non-leaf
> + * nodes on the current level. When non-zero, it
> + * indicates there is at least one non-leaf node on this
> + * level with dual rate private shaper support, which
> + * may not be the case for all the non-leaf nodes on
> + * this level.
> + */
> + int shaper_private_dual_rate_supported;
> +
> + /**< Minimum committed/peak rate (bytes per second) for
> + * private shapers of the non-leaf nodes of this level.
> + * Valid only when private shaper is supported on this
> + * level.
> + */
> + uint64_t shaper_private_rate_min;
> +
> + /**< Maximum committed/peak rate (bytes per second) for
> + * private shapers of the non-leaf nodes on this level.
> + * Valid only when private shaper is supported on this
> + * level.
> + */
> + uint64_t shaper_private_rate_max;
> +
> + /**< Maximum number of shared shapers that any non-leaf
> + * node on this level can be part of. The value of zero
> + * indicates that shared shapers are not supported by
> + * the non-leaf nodes on this level. When non-zero, it
> + * indicates there is at least one non-leaf node on this
> + * level that meets this condition, which may not be the
> + * case for all the non-leaf nodes on this level.
> + */
> + uint32_t shaper_shared_n_max;
> +
> + /**< Maximum number of children nodes. This parameter
> + * indicates that there is at least one non-leaf node on
> + * this level that can be configured with this many
> + * children nodes, which might not be true for all the
> + * non-leaf nodes on this level.
> + */
> + uint32_t sched_n_children_max;
> +
> + /**< Maximum number of supported priority levels. This
> + * parameter indicates that there is at least one
> + * non-leaf node on this level that can be configured
> + * with this many priority levels for managing its
> + * children nodes, which might not be true for all the
> + * non-leaf nodes on this level. The value of zero is
> + * invalid. The value of 1 indicates that only priority
> + * 0 is supported, which essentially means that Strict
> + * Priority (SP) algorithm is not supported on this
> + * level.
> + */
> + uint32_t sched_sp_n_priorities_max;
> +
> + /**< Maximum number of sibling nodes that can have the
> + * same priority at any given time, i.e. maximum size of
> + * the WFQ sibling node group. This parameter indicates
> + * there is at least one non-leaf node on this level
> + * that meets this condition, which may not be true for
> + * all the non-leaf nodes on this level. The value of
> + * zero is invalid. The value of 1 indicates that WFQ
> + * algorithm is not supported on this level. The maximum
> + * value is *sched_n_children_max*.
> + */
> + uint32_t sched_wfq_n_children_per_group_max;
> +
> + /**< Maximum number of priority levels that can have
> + * more than one child node at any given time, i.e.
> + * maximum number of WFQ sibling node groups that
> + * have two or more members. This parameter indicates
> + * there is at least one non-leaf node on this level
> + * that meets this condition, which might not be true
> + * for all the non-leaf nodes. The value of zero states
> + * that WFQ algorithm is not supported on this level.
> + * The value of 1 indicates that
> + * (*sched_sp_n_priorities_max* - 1) priority levels on
> + * this level have at most one child node, so there can
> + * be only one priority level with two or more sibling
> + * nodes making up a WFQ group on this level. The
> + * maximum value is:
> + * min(floor(*sched_n_children_max* / 2),
> + * *sched_sp_n_priorities_max*).
> + */
> + uint32_t sched_wfq_n_groups_max;
> +
> + /**< Maximum WFQ weight. The value of 1 indicates that
> + * all sibling nodes on this level with same priority
> + * have the same WFQ weight, so on this level WFQ is
> + * reduced to FQ.
> + */
> + uint32_t sched_wfq_weight_max;
> +
> + /**< Mask of statistics counter types supported by the
> + * non-leaf nodes on this level. Every supported
> + * statistics counter type is supported by at least one
> + * non-leaf node on this level, which may not be true
> + * for all the non-leaf nodes on this level.
> + * @see enum rte_tm_stats_type
> + */
> + uint64_t stats_mask;
> + } nonleaf;
> +
> + /**< Items valid only for the leaf nodes on this level. */
> + struct {
> + /**< Private shaper support. When non-zero, it indicates
> + * there is at least one leaf node on this level with
> + * private shaper support, which may not be the case for
> + * all the leaf nodes on this level.
> + */
> + int shaper_private_supported;
> +
> + /**< Dual rate support for private shaper. Valid only
> + * when private shaper is supported for the leaf nodes
> + * on this level. When non-zero, it indicates there is
> + * at least one leaf node on this level with dual rate
> + * private shaper support, which may not be the case for
> + * all the leaf nodes on this level.
> + */
> + int shaper_private_dual_rate_supported;
> +
> + /**< Minimum committed/peak rate (bytes per second) for
> + * private shapers of the leaf nodes of this level.
> + * Valid only when private shaper is supported for the
> + * leaf nodes on this level.
> + */
> + uint64_t shaper_private_rate_min;
> +
> + /**< Maximum committed/peak rate (bytes per second) for
> + * private shapers of the leaf nodes on this level.
> + * Valid only when private shaper is supported for the
> + * leaf nodes on this level.
> + */
> + uint64_t shaper_private_rate_max;
> +
> + /**< Maximum number of shared shapers that any leaf node
> + * on this level can be part of. The value of zero
> + * indicates that shared shapers are not supported by
> + * the leaf nodes on this level. When non-zero, it
> + * indicates there is at least one leaf node on this
> + * level that meets this condition, which may not be the
> + * case for all the leaf nodes on this level.
> + */
> + uint32_t shaper_shared_n_max;
> +
> + /**< Head drop algorithm support. When non-zero, this
> + * parameter indicates that there is at least one leaf
> + * node on this level that supports the head drop
> + * algorithm, which might not be true for all the leaf
> + * nodes on this level.
> + */
> + int cman_head_drop_supported;
> +
> + /**< Private WRED context support. When non-zero, it
> + * indicates there is at least one node on this level
> + * with private WRED context support, which may not be
> + * true for all the leaf nodes on this level. */
> + int cman_wred_context_private_supported;
> +
> + /**< Maximum number of shared WRED contexts that any
> + * leaf node on this level can be part of. The value of
> + * zero indicates that shared WRED contexts are not
> + * supported by the leaf nodes on this level. When
> + * non-zero, it indicates there is at least one leaf
> + * node on this level that meets this condition, which
> + * may not be the case for all the leaf nodes on this
> + * level.
> + */
> + uint32_t cman_wred_context_shared_n_max;
> +
> + /**< Mask of statistics counter types supported by the
> + * leaf nodes on this level. Every supported statistics
> + * counter type is supported by at least one leaf node
> + * on this level, which may not be true for all the leaf
> + * nodes on this level.
> + * @see enum rte_tm_stats_type
> + */
> + uint64_t stats_mask;
> + } leaf;
> + };
> +};
> +
> +/**
> + * Traffic manager node capabilities
> + */
> +struct rte_tm_node_capabilities {
> + /**< Private shaper support for the current node. */
> + int shaper_private_supported;
> +
> + /**< Dual rate shaping support for private shaper of current node.
> + * Valid only when private shaper is supported by the current node.
> + */
> + int shaper_private_dual_rate_supported;
> +
> + /**< Minimum committed/peak rate (bytes per second) for private
> + * shaper of current node. Valid only when private shaper is supported
> + * by the current node.
> + */
> + uint64_t shaper_private_rate_min;
> +
> + /**< Maximum committed/peak rate (bytes per second) for private
> + * shaper of current node. Valid only when private shaper is supported
> + * by the current node.
> + */
> + uint64_t shaper_private_rate_max;
> +
> + /**< Maximum number of shared shapers the current node can be part of.
> + * The value of zero indicates that shared shapers are not supported by
> + * the current node.
> + */
> + uint32_t shaper_shared_n_max;
> +
> + union {
> + /**< Items valid only for non-leaf nodes. */
> + struct {
> + /**< Maximum number of children nodes. */
> + uint32_t sched_n_children_max;
> +
> + /**< Maximum number of supported priority levels. The
> + * value of zero is invalid. The value of 1 indicates
> + * that only priority 0 is supported, which essentially
> + * means that Strict Priority (SP) algorithm is not
> + * supported.
> + */
> + uint32_t sched_sp_n_priorities_max;
> +
> + /**< Maximum number of sibling nodes that can have the
> + * same priority at any given time, i.e. maximum size
> + * of the WFQ sibling node group. The value of zero
> + * is invalid. The value of 1 indicates that WFQ
> + * algorithm is not supported. The maximum value is
> + * *sched_n_children_max*.
> + */
> + uint32_t sched_wfq_n_children_per_group_max;
> +
> + /**< Maximum number of priority levels that can have
> + * more than one child node at any given time, i.e.
> + * maximum number of WFQ sibling node groups that have
> + * two or more members. The value of zero states that
> + * WFQ algorithm is not supported. The value of 1
> + * indicates that (*sched_sp_n_priorities_max* - 1)
> + * priority levels have at most one child node, so there
> + * can be only one priority level with two or more
> + * sibling nodes making up a WFQ group. The maximum
> + * value is: min(floor(*sched_n_children_max* / 2),
> + * *sched_sp_n_priorities_max*).
> + */
> + uint32_t sched_wfq_n_groups_max;
> +
> + /**< Maximum WFQ weight. The value of 1 indicates that
> + * all sibling nodes with same priority have the same
> + * WFQ weight, so WFQ is reduced to FQ.
> + */
> + uint32_t sched_wfq_weight_max;
> + } nonleaf;
> +
> + /**< Items valid only for leaf nodes. */
> + struct {
> + /**< Head drop algorithm support for current node. */
> + int cman_head_drop_supported;
> +
> + /**< Private WRED context support for current node. */
> + int cman_wred_context_private_supported;
> +
> + /**< Maximum number of shared WRED contexts the current
> + * node can be part of. The value of zero indicates that
> + * shared WRED contexts are not supported by the current
> + * node.
> + */
> + uint32_t cman_wred_context_shared_n_max;
> + } leaf;
> + };
> +
> + /**< Mask of statistics counter types supported by the current node.
> + * @see enum rte_tm_stats_type
> + */
> + uint64_t stats_mask;
> +};
> +
> +/**
> + * Congestion management (CMAN) mode
> + *
> + * This is used for controlling the admission of packets into a packet queue or
> + * group of packet queues on congestion. On request of writing a new packet
> + * into the current queue while the queue is full, the *tail drop* algorithm
> + * drops the new packet while leaving the queue unmodified, as opposed to *head
> + * drop* algorithm, which drops the packet at the head of the queue (the oldest
> + * packet waiting in the queue) and admits the new packet at the tail of the
> + * queue.
> + *
> + * The *Random Early Detection (RED)* algorithm works by proactively dropping
> + * more and more input packets as the queue occupancy builds up. When the queue
> + * is full or almost full, RED effectively works as *tail drop*. The *Weighted
> + * RED* algorithm uses a separate set of RED thresholds for each packet color.
> + */
> +enum rte_tm_cman_mode {
> + RTE_TM_CMAN_TAIL_DROP = 0, /**< Tail drop */
> + RTE_TM_CMAN_HEAD_DROP, /**< Head drop */
> + RTE_TM_CMAN_WRED, /**< Weighted Random Early Detection (WRED) */
> +};
> +
> +/**
> + * Random Early Detection (RED) profile
> + */
> +struct rte_tm_red_params {
> + /**< Minimum queue threshold */
> + uint16_t min_th;
> +
> + /**< Maximum queue threshold */
> + uint16_t max_th;
> +
> + /**< Inverse of packet marking probability maximum value (maxp), i.e.
> + * maxp_inv = 1 / maxp
> + */
> + uint16_t maxp_inv;
> +
> + /**< Negated log2 of queue weight (wq), i.e. wq = 1 / (2 ^ wq_log2) */
> + uint16_t wq_log2;
> +};
> +
> +/**
> + * Weighted RED (WRED) profile
> + *
> + * Multiple WRED contexts can share the same WRED profile. Each leaf node with
> + * WRED enabled as its congestion management mode has zero or one private WRED
> + * context (only one leaf node using it) and/or zero, one or several shared
> + * WRED contexts (multiple leaf nodes use the same WRED context). A private
> + * WRED context is used to perform congestion management for a single leaf
> + * node, while a shared WRED context is used to perform congestion management
> + * for a group of leaf nodes.
> + */
> +struct rte_tm_wred_params {
> + /**< One set of RED parameters per packet color */
> + struct rte_tm_red_params red_params[RTE_TM_COLORS];
> +};
> +
> +/**
> + * Token bucket
> + */
> +struct rte_tm_token_bucket {
> + /**< Token bucket rate (bytes per second) */
> + uint64_t rate;
> +
> + /**< Token bucket size (bytes), a.k.a. max burst size */
> + uint64_t size;
> +};
> +
> +/**
> + * Shaper (rate limiter) profile
> + *
> + * Multiple shaper instances can share the same shaper profile. Each node has
> + * zero or one private shaper (only one node using it) and/or zero, one or
> + * several shared shapers (multiple nodes use the same shaper instance).
> + * A private shaper is used to perform traffic shaping for a single node, while
> + * a shared shaper is used to perform traffic shaping for a group of nodes.
> + *
> + * Single rate shapers use a single token bucket. A single rate shaper can be
> + * configured by setting the rate of the committed bucket to zero, which
> + * effectively disables this bucket. The peak bucket is used to limit the rate
> + * and the burst size for the current shaper.
> + *
> + * Dual rate shapers use both the committed and the peak token buckets. The
> + * rate of the peak bucket has to be bigger than zero, as well as greater than
> + * or equal to the rate of the committed bucket.
> + */
> +struct rte_tm_shaper_params {
> + /**< Committed token bucket */
> + struct rte_tm_token_bucket committed;
> +
> + /**< Peak token bucket */
> + struct rte_tm_token_bucket peak;
> +
> + /**< Signed value to be added to the length of each packet for the
> + * purpose of shaping. Can be used to correct the packet length with
> + * the framing overhead bytes that are also consumed on the wire (e.g.
> + * RTE_TM_ETH_FRAMING_OVERHEAD_FCS).
> + */
> + int32_t pkt_length_adjust;
> +};
> +
> +/**
> + * Node parameters
> + *
> + * Each non-leaf node has multiple inputs (its children nodes) and single output
> + * (which is input to its parent node). It arbitrates its inputs using Strict
> + * Priority (SP) and Weighted Fair Queuing (WFQ) algorithms to schedule input
> + * packets to its output while observing its shaping (rate limiting)
> + * constraints.
> + *
> + * Algorithms such as Weighted Round Robin (WRR), Byte-level WRR, Deficit WRR
> + * (DWRR), etc. are considered approximations of the WFQ ideal and are
> + * assimilated to WFQ, although an associated implementation-dependent trade-off
> + * on accuracy, performance and resource usage might exist.
> + *
> + * Children nodes with different priorities are scheduled using the SP algorithm
> + * based on their priority, with zero (0) as the highest priority. Children with
> + * the same priority are scheduled using the WFQ algorithm according to their
> + * weights. The WFQ weight of a given child node is relative to the sum of the
> + * weights of all its sibling nodes that have the same priority, with one (1) as
> + * the lowest weight. For each SP priority, the WFQ weight mode can be set as
> + * either byte-based or packet-based.
> + *
> + * Each leaf node sits on top of a TX queue of the current Ethernet port. Hence,
> + * the leaf nodes are predefined, with their node IDs set to 0 .. (N-1), where N
> + * is the number of TX queues configured for the current Ethernet port. The
> + * non-leaf nodes have their IDs generated by the application.
> + */
> +struct rte_tm_node_params {
> + /**< Shaper profile for the private shaper. The absence of the private
> + * shaper for the current node is indicated by setting this parameter
> + * to RTE_TM_SHAPER_PROFILE_ID_NONE.
> + */
> + uint32_t shaper_profile_id;
> +
> + /**< User allocated array of valid shared shaper IDs. */
> + uint32_t *shared_shaper_id;
> +
> + /**< Number of shared shaper IDs in the *shared_shaper_id* array. */
> + uint32_t n_shared_shapers;
> +
> + union {
> + /**< Parameters only valid for non-leaf nodes. */
> + struct {
> + /**< WFQ weight mode for each SP priority. When NULL, it
> + * indicates that WFQ is to be used for all priorities.
> + * When non-NULL, it points to a pre-allocated array of
> + * *n_sp_priorities* values, with non-zero value for
> + * byte-mode and zero for packet-mode.
> + */
> + int *wfq_weight_mode;
> +
> + /**< Number of SP priorities. */
> + uint32_t n_sp_priorities;
> + } nonleaf;
> +
> + /**< Parameters only valid for leaf nodes. */
> + struct {
> + /**< Congestion management mode */
> + enum rte_tm_cman_mode cman;
> +
> + /**< WRED parameters (only valid when *cman* is set to
> + * WRED).
> + */
> + struct {
> + /**< WRED profile for private WRED context. The
> + * absence of a private WRED context for the
> + * current leaf node is indicated by value
> + * RTE_TM_WRED_PROFILE_ID_NONE.
> + */
> + uint32_t wred_profile_id;
> +
> + /**< User allocated array of shared WRED context
> + * IDs. When set to NULL, it indicates that the
> + * current leaf node should not currently be
> + * part of any shared WRED contexts.
> + */
> + uint32_t *shared_wred_context_id;
> +
> + /**< Number of elements in the
> + * *shared_wred_context_id* array. Only valid
> + * when *shared_wred_context_id* is non-NULL,
> + * in which case it should be non-zero.
> + */
> + uint32_t n_shared_wred_contexts;
> + } wred;
> + } leaf;
> + };
> +
> + /**< Mask of statistics counter types to be enabled for this node. This
> + * needs to be a subset of the statistics counter types available for
> + * the current node. Any statistics counter type not included in this
> + * set is to be disabled for the current node.
> + * @see enum rte_tm_stats_type
> + */
> + uint64_t stats_mask;
> +};
> +
> +/**
> + * Verbose error types.
> + *
> + * Most of them provide the type of the object referenced by struct
> + * rte_tm_error::cause.
> + */
> +enum rte_tm_error_type {
> + RTE_TM_ERROR_TYPE_NONE, /**< No error. */
> + RTE_TM_ERROR_TYPE_UNSPECIFIED, /**< Cause unspecified. */
> + RTE_TM_ERROR_TYPE_CAPABILITIES,
> + RTE_TM_ERROR_TYPE_LEVEL_ID,
> + RTE_TM_ERROR_TYPE_WRED_PROFILE,
> + RTE_TM_ERROR_TYPE_WRED_PROFILE_GREEN,
> + RTE_TM_ERROR_TYPE_WRED_PROFILE_YELLOW,
> + RTE_TM_ERROR_TYPE_WRED_PROFILE_RED,
> + RTE_TM_ERROR_TYPE_WRED_PROFILE_ID,
> + RTE_TM_ERROR_TYPE_SHARED_WRED_CONTEXT_ID,
> + RTE_TM_ERROR_TYPE_SHAPER_PROFILE,
> + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_COMMITTED_RATE,
> + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_COMMITTED_SIZE,
> + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_PEAK_RATE,
> + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_PEAK_SIZE,
> + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_PKT_ADJUST_LEN,
> + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_ID,
> + RTE_TM_ERROR_TYPE_SHARED_SHAPER_ID,
> + RTE_TM_ERROR_TYPE_NODE_PARENT_NODE_ID,
> + RTE_TM_ERROR_TYPE_NODE_PRIORITY,
> + RTE_TM_ERROR_TYPE_NODE_WEIGHT,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_SHAPER_PROFILE_ID,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_SHARED_SHAPER_ID,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_N_SHARED_SHAPERS,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_WFQ_WEIGHT_MODE,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_N_SP_PRIORITIES,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_CMAN,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_WRED_PROFILE_ID,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_SHARED_WRED_CONTEXT_ID,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_N_SHARED_WRED_CONTEXTS,
> + RTE_TM_ERROR_TYPE_NODE_PARAMS_STATS,
> + RTE_TM_ERROR_TYPE_NODE_ID,
> +};
> +
> +/**
> + * Verbose error structure definition.
> + *
> + * This object is normally allocated by applications and set by PMDs, the
> + * message points to a constant string which does not need to be freed by
> + * the application, however its pointer can be considered valid only as long
> + * as its associated DPDK port remains configured. Closing the underlying
> + * device or unloading the PMD invalidates it.
> + *
> + * Both cause and message may be NULL regardless of the error type.
> + */
> +struct rte_tm_error {
> + enum rte_tm_error_type type; /**< Cause field and error type. */
> + const void *cause; /**< Object responsible for the error. */
> + const char *message; /**< Human-readable error message. */
> +};
> +
> +/**
> + * Traffic manager get number of leaf nodes
> + *
> + * Each leaf node sits on on top of a TX queue of the current Ethernet port.
> + * Therefore, the set of leaf nodes is predefined, their number is always equal
> + * to N (where N is the number of TX queues configured for the current port)
> + * and their IDs are 0 .. (N-1).
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[out] n_leaf_nodes
> + * Number of leaf nodes for the current port.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_get_number_of_leaf_nodes(uint8_t port_id,
> + uint32_t *n_leaf_nodes,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node type (i.e. leaf or non-leaf) get
> + *
> + * The leaf nodes have predefined IDs in the range of 0 .. (N-1), where N is
> + * the number of TX queues of the current Ethernet port. The non-leaf nodes
> + * have their IDs generated by the application outside of the above range,
> + * which is reserved for leaf nodes.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID value. Needs to be valid.
> + * @param[out] is_leaf
> + * Set to non-zero value when node is leaf and to zero otherwise (non-leaf).
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_node_type_get(uint8_t port_id,
> + uint32_t node_id,
> + int *is_leaf,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node level get
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID value. Needs to be valid.
> + * @param[out] level_id
> + * Node level ID. Needs to be non-NULL.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_node_level_get(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t *level_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager capabilities get
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[out] cap
> + * Traffic manager capabilities. Needs to be pre-allocated and valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_capabilities_get(uint8_t port_id,
> + struct rte_tm_capabilities *cap,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager level capabilities get
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] level_id
> + * The hierarchy level identifier. The value of 0 identifies the level of the
> + * root node.
> + * @param[out] cap
> + * Traffic manager level capabilities. Needs to be pre-allocated and valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_level_capabilities_get(uint8_t port_id,
> + uint32_t level_id,
> + struct rte_tm_level_capabilities *cap,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node capabilities get
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[out] cap
> + * Traffic manager node capabilities. Needs to be pre-allocated and valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_node_capabilities_get(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_node_capabilities *cap,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager WRED profile add
> + *
> + * Create a new WRED profile with ID set to *wred_profile_id*. The new profile
> + * is used to create one or several WRED contexts.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] wred_profile_id
> + * WRED profile ID for the new profile. Needs to be unused.
> + * @param[in] profile
> + * WRED profile parameters. Needs to be pre-allocated and valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_wred_profile_add(uint8_t port_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_wred_params *profile,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager WRED profile delete
> + *
> + * Delete an existing WRED profile. This operation fails when there is
> + * currently at least one user (i.e. WRED context) of this WRED profile.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] wred_profile_id
> + * WRED profile ID. Needs to be the valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_wred_profile_delete(uint8_t port_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager shared WRED context add or update
> + *
> + * When *shared_wred_context_id* is invalid, a new WRED context with this ID is
> + * created by using the WRED profile identified by *wred_profile_id*.
> + *
> + * When *shared_wred_context_id* is valid, this WRED context is no longer using
> + * the profile previously assigned to it and is updated to use the profile
> + * identified by *wred_profile_id*.
> + *
> + * A valid shared WRED context can be assigned to several hierarchy leaf nodes
> + * configured to use WRED as the congestion management mode.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] shared_wred_context_id
> + * Shared WRED context ID
> + * @param[in] wred_profile_id
> + * WRED profile ID. Needs to be the valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_shared_wred_context_add_update(uint8_t port_id,
> + uint32_t shared_wred_context_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager shared WRED context delete
> + *
> + * Delete an existing shared WRED context. This operation fails when there is
> + * currently at least one user (i.e. hierarchy leaf node) of this shared WRED
> + * context.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] shared_wred_context_id
> + * Shared WRED context ID. Needs to be the valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_shared_wred_context_delete(uint8_t port_id,
> + uint32_t shared_wred_context_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager shaper profile add
> + *
> + * Create a new shaper profile with ID set to *shaper_profile_id*. The new
> + * shaper profile is used to create one or several shapers.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] shaper_profile_id
> + * Shaper profile ID for the new profile. Needs to be unused.
> + * @param[in] profile
> + * Shaper profile parameters. Needs to be pre-allocated and valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_shaper_profile_add(uint8_t port_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_shaper_params *profile,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager shaper profile delete
> + *
> + * Delete an existing shaper profile. This operation fails when there is
> + * currently at least one user (i.e. shaper) of this shaper profile.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] shaper_profile_id
> + * Shaper profile ID. Needs to be the valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_shaper_profile_delete(uint8_t port_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager shared shaper add or update
> + *
> + * When *shared_shaper_id* is not a valid shared shaper ID, a new shared shaper
> + * with this ID is created using the shaper profile identified by
> + * *shaper_profile_id*.
> + *
> + * When *shared_shaper_id* is a valid shared shaper ID, this shared shaper is
> + * no longer using the shaper profile previously assigned to it and is updated
> + * to use the shaper profile identified by *shaper_profile_id*.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] shared_shaper_id
> + * Shared shaper ID
> + * @param[in] shaper_profile_id
> + * Shaper profile ID. Needs to be the valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_shared_shaper_add_update(uint8_t port_id,
> + uint32_t shared_shaper_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager shared shaper delete
> + *
> + * Delete an existing shared shaper. This operation fails when there is
> + * currently at least one user (i.e. hierarchy node) of this shared shaper.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] shared_shaper_id
> + * Shared shaper ID. Needs to be the valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_shared_shaper_delete(uint8_t port_id,
> + uint32_t shared_shaper_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node add
> + *
> + * Create new node and connect it as child of an existing node. The new node is
> + * further identified by *node_id*, which needs to be unused by any of the
> + * existing nodes. The parent node is identified by *parent_node_id*, which
> + * needs to be the valid ID of an existing non-leaf node. The parent node is
> + * going to use the provided SP *priority* and WFQ *weight* to schedule its new
> + * child node.
> + *
> + * This function has to be called for both leaf and non-leaf nodes. In the case
> + * of leaf nodes (i.e. *node_id* is within the range of 0 .. (N-1), with N as
> + * the number of configured TX queues of the current port), the leaf node is
> + * configured rather than created (as the set of leaf nodes is predefined) and
> + * it is also connected as child of an existing node.
> + *
> + * The first node that is added becomes the root node and all the nodes that
> + * are subsequently added have to be added as descendants of the root node. The
> + * parent of the root node has to be specified as RTE_TM_NODE_ID_NULL and there
> + * can only be one node with this parent ID (i.e. the root node). Further
> + * restrictions for root node: needs to be non-leaf, its private shaper profile
> + * needs to be valid and single rate, cannot use any shared shapers.
> + *
> + * When called before rte_tm_hierarchy_commit() invocation, this function is
> + * typically used to define the initial start-up hierarchy for the port.
> + * Provided that dynamic hierarchy updates are supported by the current port (as
> + * advertised in the port capability set), this function can be also called
> + * after the rte_tm_hierarchy_commit() invocation.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be unused by any of the existing nodes.
> + * @param[in] parent_node_id
> + * Parent node ID. Needs to be the valid.
> + * @param[in] priority
> + * Node priority. The highest node priority is zero. Used by the SP algorithm
> + * running on the parent of the current node for scheduling this child node.
> + * @param[in] weight
> + * Node weight. The node weight is relative to the weight sum of all siblings
> + * that have the same priority. The lowest weight is one. Used by the WFQ
> + * algorithm running on the parent of the current node for scheduling this
> + * child node.
> + * @param[in] params
> + * Node parameters. Needs to be pre-allocated and valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see rte_tm_hierarchy_commit()
> + * @see RTE_TM_UPDATE_NODE_ADD_DELETE
> + */
> +int
> +rte_tm_node_add(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t parent_node_id,
> + uint32_t priority,
> + uint32_t weight,
> + struct rte_tm_node_params *params,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node add with node level check
> + *
> + * Simple rte_tm_node_add() wrapper that also checks the node level.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be unused by any of the existing nodes.
> + * @param[in] parent_node_id
> + * Parent node ID. Needs to be the valid.
> + * @param[in] priority
> + * Node priority. The highest node priority is zero. Used by the SP algorithm
> + * running on the parent of the current node for scheduling this child node.
> + * @param[in] weight
> + * Node weight. The node weight is relative to the weight sum of all siblings
> + * that have the same priority. The lowest weight is one. Used by the WFQ
> + * algorithm running on the parent of the current node for scheduling this
> + * child node.
> + * @param[in] level_id
> + * Level ID that should be met by this node.
> + * @param[in] params
> + * Node parameters. Needs to be pre-allocated and valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +static inline int
> +rte_tm_node_add_check_level(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t parent_node_id,
> + uint32_t priority,
> + uint32_t weight,
> + uint32_t level_id,
> + struct rte_tm_node_params *params,
> + struct rte_tm_error *error)
> +{
> + uint32_t lid;
> + int status;
> +
> + status = rte_tm_node_add(port_id, node_id,
> + parent_node_id, priority, weight, params, error);
> + if (status)
> + return status;
> +
> + status = rte_tm_node_level_get(port_id, node_id, &lid, error);
> + if (status)
> + return status;
> +
> + if (lid != level_id){
> + if (error){
> + error->type = RTE_TM_ERROR_TYPE_LEVEL_ID;
> + error->cause = NULL;
> + error->message = rte_strerror(EINVAL);
> + }
> + rte_errno = EINVAL;
> + return -EINVAL;
> + }
> +
> + return 0;
> +}
> +
> +/**
> + * Traffic manager node delete
> + *
> + * Delete an existing node. This operation fails when this node currently has
> + * at least one user (i.e. child node).
> + *
> + * When called before rte_tm_hierarchy_commit() invocation, this function is
> + * typically used to define the initial start-up hierarchy for the port.
> + * Provided that dynamic hierarchy updates are supported by the current port (as
> + * advertised in the port capability set), this function can be also called
> + * after the rte_tm_hierarchy_commit() invocation.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see RTE_TM_UPDATE_NODE_ADD_DELETE
> + */
> +int
> +rte_tm_node_delete(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node suspend
> + *
> + * Suspend an existing node. While the node is in suspended state, no packet is
> + * scheduled from this node and its descendants. The node exits the suspended
> + * state through the node resume operation.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see rte_tm_node_resume()
> + * @see RTE_TM_UPDATE_NODE_SUSPEND_RESUME
> + */
> +int
> +rte_tm_node_suspend(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node resume
> + *
> + * Resume an existing node that is currently in suspended state. The node
> + * entered the suspended state as result of a previous node suspend operation.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see rte_tm_node_suspend()
> + * @see RTE_TM_UPDATE_NODE_SUSPEND_RESUME
> + */
> +int
> +rte_tm_node_resume(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager hierarchy commit
> + *
> + * This function is called during the port initialization phase (before the
> + * Ethernet port is started) to freeze the start-up hierarchy.
> + *
> + * This function typically performs the following steps:
> + * a) It validates the start-up hierarchy that was previously defined for the
> + * current port through successive rte_tm_node_add() invocations;
> + * b) Assuming successful validation, it performs all the necessary port
> + * specific configuration operations to install the specified hierarchy on
> + * the current port, with immediate effect once the port is started.
> + *
> + * This function fails when the currently configured hierarchy is not supported
> + * by the Ethernet port, in which case the user can abort or try out another
> + * hierarchy configuration (e.g. a hierarchy with less leaf nodes), which can be
> + * build from scratch (when *clear_on_fail* is enabled) or by modifying the
> + * existing hierarchy configuration (when *clear_on_fail* is disabled).
> + *
> + * Note that this function can still fail due to other causes (e.g. not enough
> + * memory available in the system, etc), even though the specified hierarchy is
> + * supported in principle by the current port.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] clear_on_fail
> + * On function call failure, hierarchy is cleared when this parameter is
> + * non-zero and preserved when this parameter is equal to zero.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see rte_tm_node_add()
> + * @see rte_tm_node_delete()
> + */
> +int
> +rte_tm_hierarchy_commit(uint8_t port_id,
> + int clear_on_fail,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node parent update
> + *
> + * Restriction for root node: its parent cannot be changed.
> + *
> + * This function can only be called after the rte_tm_hierarchy_commit()
> + * invocation. Its success depends on the port support for this operation, as
> + * advertised through the port capability set.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[in] parent_node_id
> + * Node ID for the new parent. Needs to be valid.
> + * @param[in] priority
> + * Node priority. The highest node priority is zero. Used by the SP algorithm
> + * running on the parent of the current node for scheduling this child node.
> + * @param[in] weight
> + * Node weight. The node weight is relative to the weight sum of all siblings
> + * that have the same priority. The lowest weight is zero. Used by the WFQ
> + * algorithm running on the parent of the current node for scheduling this
> + * child node.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see RTE_TM_UPDATE_NODE_PARENT_KEEP_LEVEL
> + * @see RTE_TM_UPDATE_NODE_PARENT_CHANGE_LEVEL
> + */
> +int
> +rte_tm_node_parent_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t parent_node_id,
> + uint32_t priority,
> + uint32_t weight,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node private shaper update
> + *
> + * Restriction for the root node: its private shaper profile needs to be valid
> + * and single rate.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[in] shaper_profile_id
> + * Shaper profile ID for the private shaper of the current node. Needs to be
> + * either valid shaper profile ID or RTE_TM_SHAPER_PROFILE_ID_NONE, with
> + * the latter disabling the private shaper of the current node.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_node_shaper_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node shared shapers update
> + *
> + * Restriction for root node: cannot use any shared rate shapers.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[in] shared_shaper_id
> + * Shared shaper ID. Needs to be valid.
> + * @param[in] add
> + * Set to non-zero value to add this shared shaper to current node or to zero
> + * to delete this shared shaper from current node.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_node_shared_shaper_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t shared_shaper_id,
> + int add,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node enabled statistics counters update
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[in] stats_mask
> + * Mask of statistics counter types to be enabled for the current node. This
> + * needs to be a subset of the statistics counter types available for the
> + * current node. Any statistics counter type not included in this set is to
> + * be disabled for the current node.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see enum rte_tm_stats_type
> + * @see RTE_TM_UPDATE_NODE_STATS
> + */
> +int
> +rte_tm_node_stats_update(uint8_t port_id,
> + uint32_t node_id,
> + uint64_t stats_mask,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node WFQ weight mode update
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid leaf node ID.
> + * @param[in] wfq_weight_mode
> + * WFQ weight mode for each SP priority. When NULL, it indicates that WFQ is
> + * to be used for all priorities. When non-NULL, it points to a pre-allocated
> + * array of *n_sp_priorities* values, with non-zero value for byte-mode and
> + * zero for packet-mode.
> + * @param[in] n_sp_priorities
> + * Number of SP priorities.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see RTE_TM_UPDATE_NODE_WFQ_WEIGHT_MODE
> + * @see RTE_TM_UPDATE_NODE_N_SP_PRIORITIES
> + */
> +int
> +rte_tm_node_wfq_weight_mode_update(uint8_t port_id,
> + uint32_t node_id,
> + int *wfq_weight_mode,
> + uint32_t n_sp_priorities,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node congestion management mode update
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid leaf node ID.
> + * @param[in] cman
> + * Congestion management mode.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see RTE_TM_UPDATE_NODE_CMAN
> + */
> +int
> +rte_tm_node_cman_update(uint8_t port_id,
> + uint32_t node_id,
> + enum rte_tm_cman_mode cman,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node private WRED context update
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid leaf node ID.
> + * @param[in] wred_profile_id
> + * WRED profile ID for the private WRED context of the current node. Needs to
> + * be either valid WRED profile ID or RTE_TM_WRED_PROFILE_ID_NONE, with the
> + * latter disabling the private WRED context of the current node.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_node_wred_context_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node shared WRED context update
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid leaf node ID.
> + * @param[in] shared_wred_context_id
> + * Shared WRED context ID. Needs to be valid.
> + * @param[in] add
> + * Set to non-zero value to add this shared WRED context to current node or
> + * to zero to delete this shared WRED context from current node.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + */
> +int
> +rte_tm_node_shared_wred_context_update(uint8_t port_id,
> + uint32_t node_id,
> + uint32_t shared_wred_context_id,
> + int add,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager node statistics counters read
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] node_id
> + * Node ID. Needs to be valid.
> + * @param[out] stats
> + * When non-NULL, it contains the current value for the statistics counters
> + * enabled for the current node.
> + * @param[out] stats_mask
> + * When non-NULL, it contains the mask of statistics counter types that are
> + * currently enabled for this node, indicating which of the counters
> + * retrieved with the *stats* structure are valid.
> + * @param[in] clear
> + * When this parameter has a non-zero value, the statistics counters are
> + * cleared (i.e. set to zero) immediately after they have been read,
> + * otherwise the statistics counters are left untouched.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see enum rte_tm_stats_type
> + */
> +int
> +rte_tm_node_stats_read(uint8_t port_id,
> + uint32_t node_id,
> + struct rte_tm_node_stats *stats,
> + uint64_t *stats_mask,
> + int clear,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager packet marking - VLAN DEI (IEEE 802.1Q)
> + *
> + * IEEE 802.1p maps the traffic class to the VLAN Priority Code Point (PCP)
> + * field (3 bits), while IEEE 802.1q maps the drop priority to the VLAN Drop
> + * Eligible Indicator (DEI) field (1 bit), which was previously named Canonical
> + * Format Indicator (CFI).
> + *
> + * All VLAN frames of a given color get their DEI bit set if marking is enabled
> + * for this color; otherwise, their DEI bit is left as is (either set or not).
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] mark_green
> + * Set to non-zero value to enable marking of green packets and to zero to
> + * disable it.
> + * @param[in] mark_yellow
> + * Set to non-zero value to enable marking of yellow packets and to zero to
> + * disable it.
> + * @param[in] mark_red
> + * Set to non-zero value to enable marking of red packets and to zero to
> + * disable it.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see struct rte_tm_capabilities::mark_vlan_dei_supported
> + */
> +int
> +rte_tm_mark_vlan_dei(uint8_t port_id,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager packet marking - IPv4 / IPv6 ECN (IETF RFC 3168)
> + *
> + * IETF RFCs 2474 and 3168 reorganize the IPv4 Type of Service (TOS) field
> + * (8 bits) and the IPv6 Traffic Class (TC) field (8 bits) into Differentiated
> + * Services Codepoint (DSCP) field (6 bits) and Explicit Congestion
> + * Notification (ECN) field (2 bits). The DSCP field is typically used to
> + * encode the traffic class and/or drop priority (RFC 2597), while the ECN
> + * field is used by RFC 3168 to implement a congestion notification mechanism
> + * to be leveraged by transport layer protocols such as TCP and SCTP that have
> + * congestion control mechanisms.
> + *
> + * When congestion is experienced, as alternative to dropping the packet,
> + * routers can change the ECN field of input packets from 2'b01 or 2'b10
> + * (values indicating that source endpoint is ECN-capable) to 2'b11 (meaning
> + * that congestion is experienced). The destination endpoint can use the
> + * ECN-Echo (ECE) TCP flag to relay the congestion indication back to the
> + * source endpoint, which acknowledges it back to the destination endpoint with
> + * the Congestion Window Reduced (CWR) TCP flag.
> + *
> + * All IPv4/IPv6 packets of a given color with ECN set to 2’b01 or 2’b10
> + * carrying TCP or SCTP have their ECN set to 2’b11 if the marking feature is
> + * enabled for the current color, otherwise the ECN field is left as is.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] mark_green
> + * Set to non-zero value to enable marking of green packets and to zero to
> + * disable it.
> + * @param[in] mark_yellow
> + * Set to non-zero value to enable marking of yellow packets and to zero to
> + * disable it.
> + * @param[in] mark_red
> + * Set to non-zero value to enable marking of red packets and to zero to
> + * disable it.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see struct rte_tm_capabilities::mark_ip_ecn_tcp_supported
> + * @see struct rte_tm_capabilities::mark_ip_ecn_sctp_supported
> + */
> +int
> +rte_tm_mark_ip_ecn(uint8_t port_id,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error);
> +
> +/**
> + * Traffic manager packet marking - IPv4 / IPv6 DSCP (IETF RFC 2597)
> + *
> + * IETF RFC 2597 maps the traffic class and the drop priority to the IPv4/IPv6
> + * Differentiated Services Codepoint (DSCP) field (6 bits). Here are the DSCP
> + * values proposed by this RFC:
> + *
> + * Class 1 Class 2 Class 3 Class 4
> + * +----------+----------+----------+----------+
> + * Low Drop Prec | 001010 | 010010 | 011010 | 100010 |
> + * Medium Drop Prec | 001100 | 010100 | 011100 | 100100 |
> + * High Drop Prec | 001110 | 010110 | 011110 | 100110 |
> + * +----------+----------+----------+----------+
> + *
> + * There are 4 traffic classes (classes 1 .. 4) encoded by DSCP bits 1 and 2,
> + * as well as 3 drop priorities (low/medium/high) encoded by DSCP bits 3 and 4.
> + *
> + * All IPv4/IPv6 packets have their color marked into DSCP bits 3 and 4 as
> + * follows: green mapped to Low Drop Precedence (2’b01), yellow to Medium
> + * (2’b10) and red to High (2’b11). Marking needs to be explicitly enabled
> + * for each color; when not enabled for a given color, the DSCP field of all
> + * packets with that color is left as is.
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[in] mark_green
> + * Set to non-zero value to enable marking of green packets and to zero to
> + * disable it.
> + * @param[in] mark_yellow
> + * Set to non-zero value to enable marking of yellow packets and to zero to
> + * disable it.
> + * @param[in] mark_red
> + * Set to non-zero value to enable marking of red packets and to zero to
> + * disable it.
> + * @param[out] error
> + * Error details. Filled in only on error, when not NULL.
> + * @return
> + * 0 on success, non-zero error code otherwise.
> + *
> + * @see struct rte_tm_capabilities::mark_ip_dscp_supported
> + */
> +int
> +rte_tm_mark_ip_dscp(uint8_t port_id,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error);
> +
> +#ifdef __cplusplus
> +}
> +#endif
> +
> +#endif /* __INCLUDE_RTE_TM_H__ */
> diff --git a/lib/librte_ether/rte_tm_driver.h b/lib/librte_ether/rte_tm_driver.h
> new file mode 100644
> index 0000000..c25f102
> --- /dev/null
> +++ b/lib/librte_ether/rte_tm_driver.h
> @@ -0,0 +1,373 @@
> +/*-
> + * BSD LICENSE
> + *
> + * Copyright(c) 2017 Intel Corporation. All rights reserved.
> + * All rights reserved.
> + *
> + * Redistribution and use in source and binary forms, with or without
> + * modification, are permitted provided that the following conditions
> + * are met:
> + *
> + * * Redistributions of source code must retain the above copyright
> + * notice, this list of conditions and the following disclaimer.
> + * * Redistributions in binary form must reproduce the above copyright
> + * notice, this list of conditions and the following disclaimer in
> + * the documentation and/or other materials provided with the
> + * distribution.
> + * * Neither the name of Intel Corporation nor the names of its
> + * contributors may be used to endorse or promote products derived
> + * from this software without specific prior written permission.
> + *
> + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
> + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
> + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
> + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
> + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
> + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
> + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
> + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + */
> +
> +#ifndef __INCLUDE_RTE_TM_DRIVER_H__
> +#define __INCLUDE_RTE_TM_DRIVER_H__
> +
> +/**
> + * @file
> + * RTE Generic Traffic Manager API (Driver Side)
> + *
> + * This file provides implementation helpers for internal use by PMDs, they
> + * are not intended to be exposed to applications and are not subject to ABI
> + * versioning.
> + */
> +
> +#include <stdint.h>
> +
> +#include <rte_errno.h>
> +#include "rte_ethdev.h"
> +#include "rte_tm.h"
> +
> +#ifdef __cplusplus
> +extern "C" {
> +#endif
> +
> +typedef int (*rte_tm_node_type_get_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + int *is_leaf,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node type get */
> +
> +typedef int (*rte_tm_node_level_get_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint32_t *level_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node level get */
> +
> +typedef int (*rte_tm_capabilities_get_t)(struct rte_eth_dev *dev,
> + struct rte_tm_capabilities *cap,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager capabilities get */
> +
> +typedef int (*rte_tm_level_capabilities_get_t)(struct rte_eth_dev *dev,
> + uint32_t level_id,
> + struct rte_tm_level_capabilities *cap,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager level capabilities get */
> +
> +typedef int (*rte_tm_node_capabilities_get_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + struct rte_tm_node_capabilities *cap,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node capabilities get */
> +
> +typedef int (*rte_tm_wred_profile_add_t)(struct rte_eth_dev *dev,
> + uint32_t wred_profile_id,
> + struct rte_tm_wred_params *profile,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager WRED profile add */
> +
> +typedef int (*rte_tm_wred_profile_delete_t)(struct rte_eth_dev *dev,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager WRED profile delete */
> +
> +typedef int (*rte_tm_shared_wred_context_add_update_t)(
> + struct rte_eth_dev *dev,
> + uint32_t shared_wred_context_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager shared WRED context add */
> +
> +typedef int (*rte_tm_shared_wred_context_delete_t)(
> + struct rte_eth_dev *dev,
> + uint32_t shared_wred_context_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager shared WRED context delete */
> +
> +typedef int (*rte_tm_shaper_profile_add_t)(struct rte_eth_dev *dev,
> + uint32_t shaper_profile_id,
> + struct rte_tm_shaper_params *profile,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager shaper profile add */
> +
> +typedef int (*rte_tm_shaper_profile_delete_t)(struct rte_eth_dev *dev,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager shaper profile delete */
> +
> +typedef int (*rte_tm_shared_shaper_add_update_t)(struct rte_eth_dev *dev,
> + uint32_t shared_shaper_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager shared shaper add/update */
> +
> +typedef int (*rte_tm_shared_shaper_delete_t)(struct rte_eth_dev *dev,
> + uint32_t shared_shaper_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager shared shaper delete */
> +
> +typedef int (*rte_tm_node_add_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint32_t parent_node_id,
> + uint32_t priority,
> + uint32_t weight,
> + struct rte_tm_node_params *params,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node add */
> +
> +typedef int (*rte_tm_node_delete_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node delete */
> +
> +typedef int (*rte_tm_node_suspend_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node suspend */
> +
> +typedef int (*rte_tm_node_resume_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node resume */
> +
> +typedef int (*rte_tm_hierarchy_commit_t)(struct rte_eth_dev *dev,
> + int clear_on_fail,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager hierarchy commit */
> +
> +typedef int (*rte_tm_node_parent_update_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint32_t parent_node_id,
> + uint32_t priority,
> + uint32_t weight,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node parent update */
> +
> +typedef int (*rte_tm_node_shaper_update_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint32_t shaper_profile_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node shaper update */
> +
> +typedef int (*rte_tm_node_shared_shaper_update_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint32_t shared_shaper_id,
> + int32_t add,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node shaper update */
> +
> +typedef int (*rte_tm_node_stats_update_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint64_t stats_mask,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node stats update */
> +
> +typedef int (*rte_tm_node_wfq_weight_mode_update_t)(
> + struct rte_eth_dev *dev,
> + uint32_t node_id,
> + int *wfq_weigth_mode,
> + uint32_t n_sp_priorities,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node WFQ weight mode update */
> +
> +typedef int (*rte_tm_node_cman_update_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + enum rte_tm_cman_mode cman,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node congestion management mode update */
> +
> +typedef int (*rte_tm_node_wred_context_update_t)(
> + struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint32_t wred_profile_id,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node WRED context update */
> +
> +typedef int (*rte_tm_node_shared_wred_context_update_t)(
> + struct rte_eth_dev *dev,
> + uint32_t node_id,
> + uint32_t shared_wred_context_id,
> + int add,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager node WRED context update */
> +
> +typedef int (*rte_tm_node_stats_read_t)(struct rte_eth_dev *dev,
> + uint32_t node_id,
> + struct rte_tm_node_stats *stats,
> + uint64_t *stats_mask,
> + int clear,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager read stats counters for specific node */
> +
> +typedef int (*rte_tm_mark_vlan_dei_t)(struct rte_eth_dev *dev,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager packet marking - VLAN DEI */
> +
> +typedef int (*rte_tm_mark_ip_ecn_t)(struct rte_eth_dev *dev,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager packet marking - IPv4/IPv6 ECN */
> +
> +typedef int (*rte_tm_mark_ip_dscp_t)(struct rte_eth_dev *dev,
> + int mark_green,
> + int mark_yellow,
> + int mark_red,
> + struct rte_tm_error *error);
> +/**< @internal Traffic manager packet marking - IPv4/IPv6 DSCP */
> +
> +struct rte_tm_ops {
> + /** Traffic manager node type get */
> + rte_tm_node_type_get_t node_type_get;
> + /** Traffic manager node level get */
> + rte_tm_node_level_get_t node_level_get;
> +
> + /** Traffic manager capabilities_get */
> + rte_tm_capabilities_get_t capabilities_get;
> + /** Traffic manager level capabilities_get */
> + rte_tm_level_capabilities_get_t level_capabilities_get;
> + /** Traffic manager node capabilities get */
> + rte_tm_node_capabilities_get_t node_capabilities_get;
> +
> + /** Traffic manager WRED profile add */
> + rte_tm_wred_profile_add_t wred_profile_add;
> + /** Traffic manager WRED profile delete */
> + rte_tm_wred_profile_delete_t wred_profile_delete;
> + /** Traffic manager shared WRED context add/update */
> + rte_tm_shared_wred_context_add_update_t
> + shared_wred_context_add_update;
> + /** Traffic manager shared WRED context delete */
> + rte_tm_shared_wred_context_delete_t
> + shared_wred_context_delete;
> +
> + /** Traffic manager shaper profile add */
> + rte_tm_shaper_profile_add_t shaper_profile_add;
> + /** Traffic manager shaper profile delete */
> + rte_tm_shaper_profile_delete_t shaper_profile_delete;
> + /** Traffic manager shared shaper add/update */
> + rte_tm_shared_shaper_add_update_t shared_shaper_add_update;
> + /** Traffic manager shared shaper delete */
> + rte_tm_shared_shaper_delete_t shared_shaper_delete;
> +
> + /** Traffic manager node add */
> + rte_tm_node_add_t node_add;
> + /** Traffic manager node delete */
> + rte_tm_node_delete_t node_delete;
> + /** Traffic manager node suspend */
> + rte_tm_node_suspend_t node_suspend;
> + /** Traffic manager node resume */
> + rte_tm_node_resume_t node_resume;
> + /** Traffic manager hierarchy commit */
> + rte_tm_hierarchy_commit_t hierarchy_commit;
> +
> + /** Traffic manager node parent update */
> + rte_tm_node_parent_update_t node_parent_update;
> + /** Traffic manager node shaper update */
> + rte_tm_node_shaper_update_t node_shaper_update;
> + /** Traffic manager node shared shaper update */
> + rte_tm_node_shared_shaper_update_t node_shared_shaper_update;
> + /** Traffic manager node stats update */
> + rte_tm_node_stats_update_t node_stats_update;
> + /** Traffic manager node WFQ weight mode update */
> + rte_tm_node_wfq_weight_mode_update_t node_wfq_weight_mode_update;
> + /** Traffic manager node congestion management mode update */
> + rte_tm_node_cman_update_t node_cman_update;
> + /** Traffic manager node WRED context update */
> + rte_tm_node_wred_context_update_t node_wred_context_update;
> + /** Traffic manager node shared WRED context update */
> + rte_tm_node_shared_wred_context_update_t
> + node_shared_wred_context_update;
> + /** Traffic manager read statistics counters for current node */
> + rte_tm_node_stats_read_t node_stats_read;
> +
> + /** Traffic manager packet marking - VLAN DEI */
> + rte_tm_mark_vlan_dei_t mark_vlan_dei;
> + /** Traffic manager packet marking - IPv4/IPv6 ECN */
> + rte_tm_mark_ip_ecn_t mark_ip_ecn;
> + /** Traffic manager packet marking - IPv4/IPv6 DSCP */
> + rte_tm_mark_ip_dscp_t mark_ip_dscp;
> +};
> +
> +/**
> + * Initialize generic error structure.
> + *
> + * This function also sets rte_errno to a given value.
> + *
> + * @param[out] error
> + * Pointer to error structure (may be NULL).
> + * @param[in] code
> + * Related error code (rte_errno).
> + * @param[in] type
> + * Cause field and error type.
> + * @param[in] cause
> + * Object responsible for the error.
> + * @param[in] message
> + * Human-readable error message.
> + *
> + * @return
> + * Error code.
> + */
> +static inline int
> +rte_tm_error_set(struct rte_tm_error *error,
> + int code,
> + enum rte_tm_error_type type,
> + const void *cause,
> + const char *message)
> +{
> + if (error) {
> + *error = (struct rte_tm_error){
> + .type = type,
> + .cause = cause,
> + .message = message,
> + };
> + }
> + rte_errno = code;
> + return code;
> +}
> +
> +/**
> + * Get generic traffic manager operations structure from a port
> + *
> + * @param[in] port_id
> + * The port identifier of the Ethernet device.
> + * @param[out] error
> + * Error details
> + *
> + * @return
> + * The traffic manager operations structure associated with port_id on
> + * success, NULL otherwise.
> + */
> +const struct rte_tm_ops *
> +rte_tm_ops_get(uint8_t port_id, struct rte_tm_error *error);
> +
> +#ifdef __cplusplus
> +}
> +#endif
> +
> +#endif /* __INCLUDE_RTE_TM_DRIVER_H__ */
>
Acked-by: Hemant Agrawal <hemant.agrawal at nxp.com>
More information about the dev
mailing list