From: Jie Liu <[email protected]> Update the sxe2.ini feature sheet to accurately reflect the recently implemented hardware capabilities in the sxe2 PMD.
Signed-off-by: Jie Liu <[email protected]> --- doc/guides/nics/features/sxe2.ini | 56 ++++++++++ doc/guides/nics/sxe2.rst | 164 ++++++++++++++++++++++++++++++ 2 files changed, 220 insertions(+) diff --git a/doc/guides/nics/features/sxe2.ini b/doc/guides/nics/features/sxe2.ini index 09ba2f558c..3c1e6a8a39 100644 --- a/doc/guides/nics/features/sxe2.ini +++ b/doc/guides/nics/features/sxe2.ini @@ -7,17 +7,73 @@ ; is selected. ; [Features] +Speed capabilities = Y +Link status = Y +Link status event = Y +Rx interrupt = Y Fast mbuf free = P Free Tx mbuf on demand = Y Burst mode info = Y Queue start/stop = Y +Power mgmt address monitor = Y Buffer split on Rx = P Scattered Rx = Y +Traffic manager = Y CRC offload = Y +VLAN offload = Y +QinQ offload = P L3 checksum offload = Y L4 checksum offload = Y +Timestamp offload = P +Inner L3 checksum = P +Inner L4 checksum = P Rx descriptor status = Y Tx descriptor status = Y +MTU update = Y +TSO = P +Promiscuous mode = Y +Allmulticast mode = Y +Unicast MAC filter = Y +RSS hash = Y +RSS key update = Y +RSS reta update = Y +VLAN filter = Y +Inline crypto = Y +Packet type parsing = Y +Timesync = Y +Basic stats = Y +Extended stats = Y +FW version = Y +Module EEPROM dump = Y +Multiprocess aware = Y Linux = Y x86-32 = Y x86-64 = Y + +[rte_flow items] +eth = P +geneve = Y +gre = Y +gtpu = Y +ipv4 = Y +ipv6 = Y +ipv6_frag_ext = Y +nvgre = Y +sctp = Y +tcp = Y +udp = Y +vlan = P +vxlan = Y +vxlan_gpe = Y + +[rte_flow actions] +count = Y +drop = Y +mark = Y +passthru = Y +port_representor = Y +queue = Y +represented_port = Y +rss = Y +send_to_kernel = Y +port_id = Y diff --git a/doc/guides/nics/sxe2.rst b/doc/guides/nics/sxe2.rst index 539072b076..06541bf66e 100644 --- a/doc/guides/nics/sxe2.rst +++ b/doc/guides/nics/sxe2.rst @@ -35,3 +35,167 @@ preventing unauthorized access to random physical memory. This capability allows the PMD to coexist with kernel network interfaces which remain functional, although they stop receiving unicast packets as long as they share the same MAC address. + +Configuration +------------- + +Runtime Configuration +~~~~~~~~~~~~~~~~~~~~~ + +- ``Traffic Management Scheduling Levels`` + + The DPDK Traffic Management (rte_tm) APIs can be used to configure the Tx scheduler on the NIC. + The ``sched-layer-mode`` parameter can be used to set the number of scheduling levels + in the transmit scheduling hierarchy. + The provided value must be between 0 and 3. + If the value provided is greater than the number of levels supported by the HW, + the driver will use the hardware maximum value. + +- ``flow-duplicate-pattern`` parameter [int] + + There are two options to choose: + + - 0. Prevent insertion of flow rules with the same pattern items. + In this case, duplicate rules are rejected and error code EEXIST is returned. + + - 1. Allow insertion of flow rules with the same pattern items. + In this case, the per-rule metadata flag ``switch_pattern_dup_allow`` is set + to instruct the hardware switch engine that rules with identical pattern + items are permitted. + + This option only applies to the switch engine flow type. + For the Fnav flow engine type, duplicate rules are always rejected. + + By default, the PMD will set this value to 1. + +- ``fnav-stat-type`` parameter [int] + + This parameter controls the Fnav flow engine statistics type used + for flow rule hit counting (via ``rte_flow_query``). + + - 1: Only count the number of packets. + - 2: Only count the number of bytes. + - 3: Count both packets and bytes (default). + + Default value is 3 (count both packets and bytes). + +- ``drv-sw-stats`` parameter [int] + + This parameter controls whether per-packet software statistics + (SW stats) are collected in the Rx data path. + + Hardware packet statistic counters may be inaccurate for certain + packet types due to hardware design limitations. + When accuracy of Rx packet classification statistics is critical, + enabling this parameter allows the driver to accumulate statistics + in software as packets are received, providing an alternative + statistical path that bypasses hardware counter inaccuracies. + + - 0: Disable software statistics collection (default). + The basic port statistics (``ipackets``, ``ibytes``) are reported + from the hardware counters. + - 1: Enable software statistics collection. + Per-packet software statistics are accumulated for unicast, + multicast, broadcast, and dropped packets in the Rx data path. + + When enabled, the following extended statistics (xstats) are available: + ``rx_sw_unicast_packets``, ``rx_sw_multicast_packets``, + ``rx_sw_broadcast_packets``, ``rx_sw_drop_packets``, + and ``rx_sw_drop_bytes``. + +- ``no-sched-mode`` parameter [int] + + This parameter enables non-scheduling mode (no-sched mode). + When enabled, the transmit path bypasses the hardware scheduling module + and packets are sent directly out through the port. + This results in lower transmit latency and higher throughput, + but Traffic Management (rte_tm) APIs are not supported in this mode. + + - 0: Disable non-scheduling mode (default). + The transmit path goes through the hardware scheduling hierarchy. + Traffic Management (rte_tm) APIs can be used to configure the Tx scheduler. + - 1: Enable non-scheduling mode. + The transmit path bypasses the hardware scheduling module. + Packets are sent directly from the port at full speed without scheduling. + Traffic Management (rte_tm) APIs are not available in this mode. + +- ``rx-low-latency`` parameter [int] + + This parameter controls the interrupt throttling (ITR) interval + for Rx queue interrupts. + + When enabled, the driver sets a shorter interrupt coalescing timeout + (``SXE2_ITR_INTERVAL_LOW``, approximately 1 ??s), + reducing the time between packet arrival and interrupt delivery to the CPU. + This lowers receive latency at the cost of increased CPU interrupt rate. + + When disabled (default), the driver uses the normal interrupt throttling + interval (``SXE2_ITR_INTERVAL_NORMAL``, approximately 20 ??s), + which reduces the CPU interrupt rate at the expense of higher receive latency. + + - 0: Disable Rx low latency (default). + Normal interrupt throttling interval (~20 ??s) is used. + - 1: Enable Rx low latency. + Low interrupt throttling interval (~1 ??s) is used + for reduced receive latency. + +- ``function-flow-direct`` parameter [int] + + This parameter controls whether flow rules from different functional units + (DPDK vs kernel driver) are isolated or combined when both drivers + control the same physical port. + + When the DPDK PMD and the kernel network driver coexist on the same port, + flow rules may originate from either driver. + This parameter determines how the source VSI (Virtual Switch Interface) + of each flow rule is handled during hardware programming. + + - 0 (default): Isolate flow rules between DPDK and kernel. + When ``flow_isolated`` is enabled (``rte_flow_isolate()`` called), + kernel-side flow rules take priority and DPDK-side flow rules are suppressed. + When ``flow_isolated`` is disabled, DPDK-side flow rules take priority + and kernel-side flow rules are suppressed. + Only one functional unit's flows are active at a time. + + - 1: Allow direct flow rules from both DPDK and kernel simultaneously. + Both DPDK and kernel source VSIs are preserved in the hardware flow table. + Flow rules from both sides are programmed without isolation. + + This option only applies to FNAV and ACL flow engine types + and does not apply to PF bond devices. + +Extended Statistics (xstats) +~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +The PMD provides the following extended statistics (xstats) for detailed +monitoring of receive-side packet classification and software-level accounting. +The software statistics path is provided as a workaround for hardware +counter inaccuracies on certain packet types --- it accumulates per-packet +statistics directly in the Rx data path, ensuring that unicast, multicast, +broadcast, and drop counts reflect the actual packets processed by the driver. + +Receive Software Statistics + These counters are collected in the Rx data path when ``drv-sw-stats=1`` + is configured (see the ``drv-sw-stats`` devarg above). + When ``drv-sw-stats`` is disabled (default), these xstats report zero. + + - ``rx_sw_unicast_packets``: Number of unicast packets received. + - ``rx_sw_multicast_packets``: Number of multicast packets received. + - ``rx_sw_broadcast_packets``: Number of broadcast packets received. + - ``rx_sw_drop_packets``: Number of packets dropped in the Rx data path. + - ``rx_sw_drop_bytes``: Number of bytes dropped in the Rx data path. + + When ``drv-sw-stats`` is enabled, the basic counters ``ipackets`` and + ``ibytes`` (from ``rte_eth_stats``) also reflect the software-accumulated + packet and byte counts. Otherwise, they are reported from hardware counters. + +Fnav Flow Engine Statistics + The Fnav flow engine statistics type is controlled by the ``fnav-stat-type`` + devarg (see above). Depending on the configuration: + + - ``fnav-stat-type=1``: Only packet count is available. + - ``fnav-stat-type=2``: Only byte count is available. + - ``fnav-stat-type=3`` (default): Both packet and byte counts are available. + + Flow query results (via ``rte_flow_query``) expose these per-flow counters + through the query API, not via xstats. -- 2.52.0
