1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright 2015 6WIND S.A.
3 Copyright 2015 Mellanox Technologies, Ltd
8 The MLX5 poll mode driver library (**librte_pmd_mlx5**) provides support
9 for **Mellanox ConnectX-4**, **Mellanox ConnectX-4 Lx** , **Mellanox
10 ConnectX-5**, **Mellanox ConnectX-6** and **Mellanox BlueField** families
11 of 10/25/40/50/100/200 Gb/s adapters as well as their virtual functions (VF)
14 Information and documentation about these adapters can be found on the
15 `Mellanox website <http://www.mellanox.com>`__. Help is also provided by the
16 `Mellanox community <http://community.mellanox.com/welcome>`__.
18 There is also a `section dedicated to this poll mode driver
19 <http://www.mellanox.com/page/products_dyn?product_family=209&mtag=pmd_for_dpdk>`__.
23 Due to external dependencies, this driver is disabled by default. It must
24 be enabled manually by setting ``CONFIG_RTE_LIBRTE_MLX5_PMD=y`` and
27 Implementation details
28 ----------------------
30 Besides its dependency on libibverbs (that implies libmlx5 and associated
31 kernel support), librte_pmd_mlx5 relies heavily on system calls for control
32 operations such as querying/updating the MTU and flow control parameters.
34 For security reasons and robustness, this driver only deals with virtual
35 memory addresses. The way resources allocations are handled by the kernel
36 combined with hardware specifications that allow it to handle virtual memory
37 addresses directly ensure that DPDK applications cannot access random
38 physical memory (or memory that does not belong to the current process).
40 This capability allows the PMD to coexist with kernel network interfaces
41 which remain functional, although they stop receiving unicast packets as
42 long as they share the same MAC address.
43 This means legacy linux control tools (for example: ethtool, ifconfig and
44 more) can operate on the same network interfaces that owned by the DPDK
47 Enabling librte_pmd_mlx5 causes DPDK applications to be linked against
53 - Multi arch support: x86_64, POWER8, ARMv8, i686.
54 - Multiple TX and RX queues.
55 - Support for scattered TX and RX frames.
56 - IPv4, IPv6, TCPv4, TCPv6, UDPv4 and UDPv6 RSS on any number of queues.
57 - Several RSS hash keys, one for each flow type.
58 - Default RSS operation with no hash key specification.
59 - Configurable RETA table.
60 - Support for multiple MAC addresses.
64 - RX CRC stripping configuration.
65 - Promiscuous mode on PF and VF.
66 - Multicast promiscuous mode on PF and VF.
67 - Hardware checksum offloads.
68 - Flow director (RTE_FDIR_MODE_PERFECT, RTE_FDIR_MODE_PERFECT_MAC_VLAN and
70 - Flow API, including :ref:`flow_isolated_mode`.
72 - KVM and VMware ESX SR-IOV modes are supported.
73 - RSS hash result is supported.
74 - Hardware TSO for generic IP or UDP tunnel, including VXLAN and GRE.
75 - Hardware checksum Tx offload for generic IP or UDP tunnel, including VXLAN and GRE.
77 - Statistics query including Basic, Extended and per queue.
79 - Tunnel types: VXLAN, L3 VXLAN, VXLAN-GPE, GRE, MPLSoGRE, MPLSoUDP, IP-in-IP.
80 - Tunnel HW offloads: packet type, inner/outer RSS, IP and UDP checksum verification.
81 - NIC HW offloads: encapsulation (vxlan, gre, mplsoudp, mplsogre), NAT, routing, TTL
82 increment/decrement, count, drop, mark. For details please see :ref:`Supported hardware offloads using rte_flow API`.
83 - Flow insertion rate of more then million flows per second, when using Direct Rules.
84 - Support for multiple rte_flow groups.
90 - For secondary process:
92 - Forked secondary process not supported.
93 - External memory unregistered in EAL memseg list cannot be used for DMA
94 unless such memory has been registered by ``mlx5_mr_update_ext_mp()`` in
95 primary process and remapped to the same virtual address in secondary
96 process. If the external memory is registered by primary process but has
97 different virtual address in secondary process, unexpected error may happen.
99 - Flow pattern without any specific vlan will match for vlan packets as well:
101 When VLAN spec is not specified in the pattern, the matching rule will be created with VLAN as a wild card.
102 Meaning, the flow rule::
104 flow create 0 ingress pattern eth / vlan vid is 3 / ipv4 / end ...
106 Will only match vlan packets with vid=3. and the flow rules::
108 flow create 0 ingress pattern eth / ipv4 / end ...
112 flow create 0 ingress pattern eth / vlan / ipv4 / end ...
114 Will match any ipv4 packet (VLAN included).
116 - A multi segment packet must have less than 6 segments in case the Tx burst function
117 is set to multi-packet send or Enhanced multi-packet send. Otherwise it must have
118 less than 50 segments.
120 - Flows with a VXLAN Network Identifier equal (or ends to be equal)
121 to 0 are not supported.
123 - VXLAN TSO and checksum offloads are not supported on VM.
125 - L3 VXLAN and VXLAN-GPE tunnels cannot be supported together with MPLSoGRE and MPLSoUDP.
127 - VF: flow rules created on VF devices can only match traffic targeted at the
128 configured MAC addresses (see ``rte_eth_dev_mac_addr_add()``).
132 MAC addresses not already present in the bridge table of the associated
133 kernel network device will be added and cleaned up by the PMD when closing
134 the device. In case of ungraceful program termination, some entries may
135 remain present and should be removed manually by other means.
137 - When Multi-Packet Rx queue is configured (``mprq_en``), a Rx packet can be
138 externally attached to a user-provided mbuf with having EXT_ATTACHED_MBUF in
139 ol_flags. As the mempool for the external buffer is managed by PMD, all the
140 Rx mbufs must be freed before the device is closed. Otherwise, the mempool of
141 the external buffers will be freed by PMD and the application which still
142 holds the external buffers may be corrupted.
144 - If Multi-Packet Rx queue is configured (``mprq_en``) and Rx CQE compression is
145 enabled (``rxq_cqe_comp_en``) at the same time, RSS hash result is not fully
146 supported. Some Rx packets may not have PKT_RX_RSS_HASH.
148 - IPv6 Multicast messages are not supported on VM, while promiscuous mode
149 and allmulticast mode are both set to off.
150 To receive IPv6 Multicast messages on VM, explicitly set the relevant
151 MAC address using rte_eth_dev_mac_addr_add() API.
153 - E-Switch decapsulation Flow:
155 - can be applied to PF port only.
156 - must specify VF port action (packet redirection from PF to VF).
157 - optionally may specify tunnel inner source and destination MAC addresses.
159 - E-Switch encapsulation Flow:
161 - can be applied to VF ports only.
162 - must specify PF port action (packet redirection from VF to PF).
164 - ICMP/ICMP6 code/type matching cannot be supported togeter with IP-in-IP tunnel.
168 - No mbuf headroom space is created for RX packets when LRO is configured.
169 - ``scatter_fcs`` is disabled when LRO is configured.
174 MLX5 supports various of methods to report statistics:
176 Port statistics can be queried using ``rte_eth_stats_get()``. The received and sent statistics are through SW only and counts the number of packets received or sent successfully by the PMD. The imissed counter is the amount of packets that could not be delivered to SW because a queue was full. Packets not received due to congestion in the bus or on the NIC can be queried via the rx_discards_phy xstats counter.
178 Extended statistics can be queried using ``rte_eth_xstats_get()``. The extended statistics expose a wider set of counters counted by the device. The extended port statistics counts the number of packets received or sent successfully by the port. As Mellanox NICs are using the :ref:`Bifurcated Linux Driver <linux_gsg_linux_drivers>` those counters counts also packet received or sent by the Linux kernel. The counters with ``_phy`` suffix counts the total events on the physical port, therefore not valid for VF.
180 Finally per-flow statistics can by queried using ``rte_flow_query`` when attaching a count action for specific flow. The flow counter counts the number of packets received successfully by the port and match the specific flow.
188 These options can be modified in the ``.config`` file.
190 - ``CONFIG_RTE_LIBRTE_MLX5_PMD`` (default **n**)
192 Toggle compilation of librte_pmd_mlx5 itself.
194 - ``CONFIG_RTE_IBVERBS_LINK_DLOPEN`` (default **n**)
196 Build PMD with additional code to make it loadable without hard
197 dependencies on **libibverbs** nor **libmlx5**, which may not be installed
198 on the target system.
200 In this mode, their presence is still required for it to run properly,
201 however their absence won't prevent a DPDK application from starting (with
202 ``CONFIG_RTE_BUILD_SHARED_LIB`` disabled) and they won't show up as
203 missing with ``ldd(1)``.
205 It works by moving these dependencies to a purpose-built rdma-core "glue"
206 plug-in which must either be installed in a directory whose name is based
207 on ``CONFIG_RTE_EAL_PMD_PATH`` suffixed with ``-glue`` if set, or in a
208 standard location for the dynamic linker (e.g. ``/lib``) if left to the
209 default empty string (``""``).
211 This option has no performance impact.
213 - ``CONFIG_RTE_IBVERBS_LINK_STATIC`` (default **n**)
215 Embed static flavor of the dependencies **libibverbs** and **libmlx5**
216 in the PMD shared library or the executable static binary.
218 - ``CONFIG_RTE_LIBRTE_MLX5_DEBUG`` (default **n**)
220 Toggle debugging code and stricter compilation flags. Enabling this option
221 adds additional run-time checks and debugging messages at the cost of
226 For BlueField, target should be set to ``arm64-bluefield-linux-gcc``. This
227 will enable ``CONFIG_RTE_LIBRTE_MLX5_PMD`` and set ``RTE_CACHE_LINE_SIZE`` to
228 64. Default armv8a configuration of make build and meson build set it to 128
229 then brings performance degradation.
231 Environment variables
232 ~~~~~~~~~~~~~~~~~~~~~
236 A list of directories in which to search for the rdma-core "glue" plug-in,
237 separated by colons or semi-colons.
239 Only matters when compiled with ``CONFIG_RTE_IBVERBS_LINK_DLOPEN``
240 enabled and most useful when ``CONFIG_RTE_EAL_PMD_PATH`` is also set,
241 since ``LD_LIBRARY_PATH`` has no effect in this case.
243 - ``MLX5_SHUT_UP_BF``
245 Configures HW Tx doorbell register as IO-mapped.
247 By default, the HW Tx doorbell is configured as a write-combining register.
248 The register would be flushed to HW usually when the write-combining buffer
249 becomes full, but it depends on CPU design.
251 Except for vectorized Tx burst routines, a write memory barrier is enforced
252 after updating the register so that the update can be immediately visible to
255 When vectorized Tx burst is called, the barrier is set only if the burst size
256 is not aligned to MLX5_VPMD_TX_MAX_BURST. However, setting this environmental
257 variable will bring better latency even though the maximum throughput can
260 Run-time configuration
261 ~~~~~~~~~~~~~~~~~~~~~~
263 - librte_pmd_mlx5 brings kernel network interfaces up during initialization
264 because it is affected by their state. Forcing them down prevents packets
267 - **ethtool** operations on related kernel interfaces also affect the PMD.
269 - ``rxq_cqe_comp_en`` parameter [int]
271 A nonzero value enables the compression of CQE on RX side. This feature
272 allows to save PCI bandwidth and improve performance. Enabled by default.
276 - x86_64 with ConnectX-4, ConnectX-4 LX, ConnectX-5, ConnectX-6 and BlueField.
277 - POWER9 and ARMv8 with ConnectX-4 LX, ConnectX-5, ConnectX-6 and BlueField.
279 - ``rxq_cqe_pad_en`` parameter [int]
281 A nonzero value enables 128B padding of CQE on RX side. The size of CQE
282 is aligned with the size of a cacheline of the core. If cacheline size is
283 128B, the CQE size is configured to be 128B even though the device writes
284 only 64B data on the cacheline. This is to avoid unnecessary cache
285 invalidation by device's two consecutive writes on to one cacheline.
286 However in some architecture, it is more beneficial to update entire
287 cacheline with padding the rest 64B rather than striding because
288 read-modify-write could drop performance a lot. On the other hand,
289 writing extra data will consume more PCIe bandwidth and could also drop
290 the maximum throughput. It is recommended to empirically set this
291 parameter. Disabled by default.
295 - CPU having 128B cacheline with ConnectX-5 and BlueField.
297 - ``rxq_pkt_pad_en`` parameter [int]
299 A nonzero value enables padding Rx packet to the size of cacheline on PCI
300 transaction. This feature would waste PCI bandwidth but could improve
301 performance by avoiding partial cacheline write which may cause costly
302 read-modify-copy in memory transaction on some architectures. Disabled by
307 - x86_64 with ConnectX-4, ConnectX-4 LX, ConnectX-5, ConnectX-6 and BlueField.
308 - POWER8 and ARMv8 with ConnectX-4 LX, ConnectX-5, ConnectX-6 and BlueField.
310 - ``mprq_en`` parameter [int]
312 A nonzero value enables configuring Multi-Packet Rx queues. Rx queue is
313 configured as Multi-Packet RQ if the total number of Rx queues is
314 ``rxqs_min_mprq`` or more and Rx scatter isn't configured. Disabled by
317 Multi-Packet Rx Queue (MPRQ a.k.a Striding RQ) can further save PCIe bandwidth
318 by posting a single large buffer for multiple packets. Instead of posting a
319 buffers per a packet, one large buffer is posted in order to receive multiple
320 packets on the buffer. A MPRQ buffer consists of multiple fixed-size strides
321 and each stride receives one packet. MPRQ can improve throughput for
322 small-packet traffic.
324 When MPRQ is enabled, max_rx_pkt_len can be larger than the size of
325 user-provided mbuf even if DEV_RX_OFFLOAD_SCATTER isn't enabled. PMD will
326 configure large stride size enough to accommodate max_rx_pkt_len as long as
327 device allows. Note that this can waste system memory compared to enabling Rx
328 scatter and multi-segment packet.
330 - ``mprq_log_stride_num`` parameter [int]
332 Log 2 of the number of strides for Multi-Packet Rx queue. Configuring more
333 strides can reduce PCIe traffic further. If configured value is not in the
334 range of device capability, the default value will be set with a warning
335 message. The default value is 4 which is 16 strides per a buffer, valid only
336 if ``mprq_en`` is set.
338 The size of Rx queue should be bigger than the number of strides.
340 - ``mprq_max_memcpy_len`` parameter [int]
342 The maximum length of packet to memcpy in case of Multi-Packet Rx queue. Rx
343 packet is mem-copied to a user-provided mbuf if the size of Rx packet is less
344 than or equal to this parameter. Otherwise, PMD will attach the Rx packet to
345 the mbuf by external buffer attachment - ``rte_pktmbuf_attach_extbuf()``.
346 A mempool for external buffers will be allocated and managed by PMD. If Rx
347 packet is externally attached, ol_flags field of the mbuf will have
348 EXT_ATTACHED_MBUF and this flag must be preserved. ``RTE_MBUF_HAS_EXTBUF()``
349 checks the flag. The default value is 128, valid only if ``mprq_en`` is set.
351 - ``rxqs_min_mprq`` parameter [int]
353 Configure Rx queues as Multi-Packet RQ if the total number of Rx queues is
354 greater or equal to this value. The default value is 12, valid only if
357 - ``txq_inline`` parameter [int]
359 Amount of data to be inlined during TX operations. This parameter is
360 deprecated and converted to the new parameter ``txq_inline_max`` providing
361 partial compatibility.
363 - ``txqs_min_inline`` parameter [int]
365 Enable inline data send only when the number of TX queues is greater or equal
368 This option should be used in combination with ``txq_inline_max`` and
369 ``txq_inline_mpw`` below and does not affect ``txq_inline_min`` settings above.
371 If this option is not specified the default value 16 is used for BlueField
372 and 8 for other platforms
374 The data inlining consumes the CPU cycles, so this option is intended to
375 auto enable inline data if we have enough Tx queues, which means we have
376 enough CPU cores and PCI bandwidth is getting more critical and CPU
377 is not supposed to be bottleneck anymore.
379 The copying data into WQE improves latency and can improve PPS performance
380 when PCI back pressure is detected and may be useful for scenarios involving
381 heavy traffic on many queues.
383 Because additional software logic is necessary to handle this mode, this
384 option should be used with care, as it may lower performance when back
385 pressure is not expected.
387 - ``txq_inline_min`` parameter [int]
389 Minimal amount of data to be inlined into WQE during Tx operations. NICs
390 may require this minimal data amount to operate correctly. The exact value
391 may depend on NIC operation mode, requested offloads, etc.
393 If ``txq_inline_min`` key is present the specified value (may be aligned
394 by the driver in order not to exceed the limits and provide better descriptor
395 space utilization) will be used by the driver and it is guaranteed the
396 requested data bytes are inlined into the WQE beside other inline settings.
397 This keys also may update ``txq_inline_max`` value (default of specified
398 explicitly in devargs) to reserve the space for inline data.
400 If ``txq_inline_min`` key is not present, the value may be queried by the
401 driver from the NIC via DevX if this feature is available. If there is no DevX
402 enabled/supported the value 18 (supposing L2 header including VLAN) is set
403 for ConnectX-4, value 58 (supposing L2-L4 headers, required by configurations
404 over E-Switch) is set for ConnectX-4 Lx, and 0 is set by default for ConnectX-5
405 and newer NICs. If packet is shorter the ``txq_inline_min`` value, the entire
408 For the ConnectX-4 and ConnectX-4 Lx NICs driver does not allow to set
409 this value below 18 (minimal L2 header, including VLAN).
411 Please, note, this minimal data inlining disengages eMPW feature (Enhanced
412 Multi-Packet Write), because last one does not support partial packet inlining.
413 This is not very critical due to minimal data inlining is mostly required
414 by ConnectX-4 and ConnectX-4 Lx, these NICs do not support eMPW feature.
416 - ``txq_inline_max`` parameter [int]
418 Specifies the maximal packet length to be completely inlined into WQE
419 Ethernet Segment for ordinary SEND method. If packet is larger than specified
420 value, the packet data won't be copied by the driver at all, data buffer
421 is addressed with a pointer. If packet length is less or equal all packet
422 data will be copied into WQE. This may improve PCI bandwidth utilization for
423 short packets significantly but requires the extra CPU cycles.
425 The data inline feature is controlled by number of Tx queues, if number of Tx
426 queues is larger than ``txqs_min_inline`` key parameter, the inline feature
427 is engaged, if there are not enough Tx queues (which means not enough CPU cores
428 and CPU resources are scarce), data inline is not performed by the driver.
429 Assigning ``txqs_min_inline`` with zero always enables the data inline.
431 The default ``txq_inline_max`` value is 290. The specified value may be adjusted
432 by the driver in order not to exceed the limit (930 bytes) and to provide better
433 WQE space filling without gaps, the adjustment is reflected in the debug log.
435 - ``txq_inline_mpw`` parameter [int]
437 Specifies the maximal packet length to be completely inlined into WQE for
438 Enhanced MPW method. If packet is large the specified value, the packet data
439 won't be copied, and data buffer is addressed with pointer. If packet length
440 is less or equal, all packet data will be copied into WQE. This may improve PCI
441 bandwidth utilization for short packets significantly but requires the extra
444 The data inline feature is controlled by number of TX queues, if number of Tx
445 queues is larger than ``txqs_min_inline`` key parameter, the inline feature
446 is engaged, if there are not enough Tx queues (which means not enough CPU cores
447 and CPU resources are scarce), data inline is not performed by the driver.
448 Assigning ``txqs_min_inline`` with zero always enables the data inline.
450 The default ``txq_inline_mpw`` value is 188. The specified value may be adjusted
451 by the driver in order not to exceed the limit (930 bytes) and to provide better
452 WQE space filling without gaps, the adjustment is reflected in the debug log.
453 Due to multiple packets may be included to the same WQE with Enhanced Multi
454 Packet Write Method and overall WQE size is limited it is not recommended to
455 specify large values for the ``txq_inline_mpw``.
457 - ``txqs_max_vec`` parameter [int]
459 Enable vectorized Tx only when the number of TX queues is less than or
460 equal to this value. This parameter is deprecated and ignored, kept
461 for compatibility issue to not prevent driver from probing.
463 - ``txq_mpw_hdr_dseg_en`` parameter [int]
465 A nonzero value enables including two pointers in the first block of TX
466 descriptor. The parameter is deprecated and ignored, kept for compatibility
469 - ``txq_max_inline_len`` parameter [int]
471 Maximum size of packet to be inlined. This limits the size of packet to
472 be inlined. If the size of a packet is larger than configured value, the
473 packet isn't inlined even though there's enough space remained in the
474 descriptor. Instead, the packet is included with pointer. This parameter
475 is deprecated and converted directly to ``txq_inline_mpw`` providing full
476 compatibility. Valid only if eMPW feature is engaged.
478 - ``txq_mpw_en`` parameter [int]
480 A nonzero value enables Enhanced Multi-Packet Write (eMPW) for ConnectX-5,
481 ConnectX-6 and BlueField. eMPW allows the TX burst function to pack up multiple
482 packets in a single descriptor session in order to save PCI bandwidth and improve
483 performance at the cost of a slightly higher CPU usage. When ``txq_inline_mpw``
484 is set along with ``txq_mpw_en``, TX burst function copies entire packet
485 data on to TX descriptor instead of including pointer of packet.
487 The Enhanced Multi-Packet Write feature is enabled by default if NIC supports
488 it, can be disabled by explicit specifying 0 value for ``txq_mpw_en`` option.
489 Also, if minimal data inlining is requested by non-zero ``txq_inline_min``
490 option or reported by the NIC, the eMPW feature is disengaged.
492 - ``tx_vec_en`` parameter [int]
494 A nonzero value enables Tx vector on ConnectX-5, ConnectX-6 and BlueField
495 NICs if the number of global Tx queues on the port is less than
496 ``txqs_max_vec``. The parameter is deprecated and ignored.
498 - ``rx_vec_en`` parameter [int]
500 A nonzero value enables Rx vector if the port is not configured in
501 multi-segment otherwise this parameter is ignored.
505 - ``vf_nl_en`` parameter [int]
507 A nonzero value enables Netlink requests from the VF to add/remove MAC
508 addresses or/and enable/disable promiscuous/all multicast on the Netdevice.
509 Otherwise the relevant configuration must be run with Linux iproute2 tools.
510 This is a prerequisite to receive this kind of traffic.
512 Enabled by default, valid only on VF devices ignored otherwise.
514 - ``l3_vxlan_en`` parameter [int]
516 A nonzero value allows L3 VXLAN and VXLAN-GPE flow creation. To enable
517 L3 VXLAN or VXLAN-GPE, users has to configure firmware and enable this
518 parameter. This is a prerequisite to receive this kind of traffic.
522 - ``dv_flow_en`` parameter [int]
524 A nonzero value enables the DV flow steering assuming it is supported
529 - ``dv_esw_en`` parameter [int]
531 A nonzero value enables E-Switch using Direct Rules.
533 Enabled by default if supported.
535 - ``mr_ext_memseg_en`` parameter [int]
537 A nonzero value enables extending memseg when registering DMA memory. If
538 enabled, the number of entries in MR (Memory Region) lookup table on datapath
539 is minimized and it benefits performance. On the other hand, it worsens memory
540 utilization because registered memory is pinned by kernel driver. Even if a
541 page in the extended chunk is freed, that doesn't become reusable until the
542 entire memory is freed.
546 - ``representor`` parameter [list]
548 This parameter can be used to instantiate DPDK Ethernet devices from
549 existing port (or VF) representors configured on the device.
551 It is a standard parameter whose format is described in
552 :ref:`ethernet_device_standard_device_arguments`.
554 For instance, to probe port representors 0 through 2::
558 - ``max_dump_files_num`` parameter [int]
560 The maximum number of files per PMD entity that may be created for debug information.
561 The files will be created in /var/log directory or in current directory.
563 set to 128 by default.
565 - ``lro_timeout_usec`` parameter [int]
567 The maximum allowed duration of an LRO session, in micro-seconds.
568 PMD will set the nearest value supported by HW, which is not bigger than
569 the input ``lro_timeout_usec`` value.
570 If this parameter is not specified, by default PMD will set
571 the smallest value supported by HW.
573 Firmware configuration
574 ~~~~~~~~~~~~~~~~~~~~~~
576 - L3 VXLAN and VXLAN-GPE destination UDP port
578 .. code-block:: console
580 mlxconfig -d <mst device> set IP_OVER_VXLAN_EN=1
581 mlxconfig -d <mst device> set IP_OVER_VXLAN_PORT=<udp dport>
583 Verify configurations are set:
585 .. code-block:: console
587 mlxconfig -d <mst device> query | grep IP_OVER_VXLAN
588 IP_OVER_VXLAN_EN True(1)
589 IP_OVER_VXLAN_PORT <udp dport>
591 - enable ICMP/ICMP6's code/type field matching
593 .. code-block:: console
595 mlxconfig -d <mst device> set FLEX_PARSER_PROFILE_ENABLE=2
597 Verify configurations are set:
599 .. code-block:: console
601 mlxconfig -d <mst device> query | grep FLEX_PARSER_PROFILE_ENABLE
602 FLEX_PARSER_PROFILE_ENABLE 2
604 - IP-in-IP tunnel enable
606 .. code-block:: console
608 mlxconfig -d <mst device> set FLEX_PARSER_PROFILE_ENABLE=0
610 Verify configurations are set:
612 .. code-block:: console
614 mlxconfig -d <mst device> query | grep FLEX_PARSER_PROFILE_ENABLE
615 FLEX_PARSER_PROFILE_ENABLE 0
620 This driver relies on external libraries and kernel drivers for resources
621 allocations and initialization. The following dependencies are not part of
622 DPDK and must be installed separately:
626 User space Verbs framework used by librte_pmd_mlx5. This library provides
627 a generic interface between the kernel and low-level user space drivers
630 It allows slow and privileged operations (context initialization, hardware
631 resources allocations) to be managed by the kernel and fast operations to
632 never leave user space.
636 Low-level user space driver library for Mellanox
637 ConnectX-4/ConnectX-5/ConnectX-6/BlueField devices, it is automatically loaded
640 This library basically implements send/receive calls to the hardware
645 They provide the kernel-side Verbs API and low level device drivers that
646 manage actual hardware initialization and resources sharing with user
649 Unlike most other PMDs, these modules must remain loaded and bound to
652 - mlx5_core: hardware driver managing Mellanox
653 ConnectX-4/ConnectX-5/ConnectX-6/BlueField devices and related Ethernet kernel
655 - mlx5_ib: InifiniBand device driver.
656 - ib_uverbs: user space driver for Verbs (entry point for libibverbs).
658 - **Firmware update**
660 Mellanox OFED/EN releases include firmware updates for
661 ConnectX-4/ConnectX-5/ConnectX-6/BlueField adapters.
663 Because each release provides new features, these updates must be applied to
664 match the kernel modules and libraries they come with.
668 Both libraries are BSD and GPL licensed. Linux kernel modules are GPL
674 Either RDMA Core library with a recent enough Linux kernel release
675 (recommended) or Mellanox OFED/EN, which provides compatibility with older
678 RDMA Core with Linux Kernel
679 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
681 - Minimal kernel version : v4.14 or the most recent 4.14-rc (see `Linux installation documentation`_)
682 - Minimal rdma-core version: v15+ commit 0c5f5765213a ("Merge pull request #227 from yishaih/tm")
683 (see `RDMA Core installation documentation`_)
684 - When building for i686 use:
686 - rdma-core version 18.0 or above built with 32bit support.
687 - Kernel version 4.14.41 or above.
689 - Starting with rdma-core v21, static libraries can be built::
692 CFLAGS=-fPIC cmake -DIN_PLACE=1 -DENABLE_STATIC=1 -GNinja ..
695 .. _`Linux installation documentation`: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/plain/Documentation/admin-guide/README.rst
696 .. _`RDMA Core installation documentation`: https://raw.githubusercontent.com/linux-rdma/rdma-core/master/README.md
698 If rdma-core libraries are built but not installed, DPDK makefile can link them,
699 thanks to these environment variables:
701 - ``EXTRA_CFLAGS=-I/path/to/rdma-core/build/include``
702 - ``EXTRA_LDFLAGS=-L/path/to/rdma-core/build/lib``
703 - ``PKG_CONFIG_PATH=/path/to/rdma-core/build/lib/pkgconfig``
708 - Mellanox OFED version: ** 4.5, 4.6** /
709 Mellanox EN version: **4.5, 4.6**
712 - ConnectX-4: **12.21.1000** and above.
713 - ConnectX-4 Lx: **14.21.1000** and above.
714 - ConnectX-5: **16.21.1000** and above.
715 - ConnectX-5 Ex: **16.21.1000** and above.
716 - ConnectX-6: **20.99.5374** and above.
717 - BlueField: **18.25.1010** and above.
719 While these libraries and kernel modules are available on OpenFabrics
720 Alliance's `website <https://www.openfabrics.org/>`__ and provided by package
721 managers on most distributions, this PMD requires Ethernet extensions that
722 may not be supported at the moment (this is a work in progress).
725 <http://www.mellanox.com/page/products_dyn?product_family=26&mtag=linux>`__ and
727 <http://www.mellanox.com/page/products_dyn?product_family=27&mtag=linux>`__
728 include the necessary support and should be used in the meantime. For DPDK,
729 only libibverbs, libmlx5, mlnx-ofed-kernel packages and firmware updates are
730 required from that distribution.
734 Several versions of Mellanox OFED/EN are available. Installing the version
735 this DPDK release was developed and tested against is strongly
736 recommended. Please check the `prerequisites`_.
741 * Mellanox(R) ConnectX(R)-4 10G MCX4111A-XCAT (1x10G)
742 * Mellanox(R) ConnectX(R)-4 10G MCX4121A-XCAT (2x10G)
743 * Mellanox(R) ConnectX(R)-4 25G MCX4111A-ACAT (1x25G)
744 * Mellanox(R) ConnectX(R)-4 25G MCX4121A-ACAT (2x25G)
745 * Mellanox(R) ConnectX(R)-4 40G MCX4131A-BCAT (1x40G)
746 * Mellanox(R) ConnectX(R)-4 40G MCX413A-BCAT (1x40G)
747 * Mellanox(R) ConnectX(R)-4 40G MCX415A-BCAT (1x40G)
748 * Mellanox(R) ConnectX(R)-4 50G MCX4131A-GCAT (1x50G)
749 * Mellanox(R) ConnectX(R)-4 50G MCX413A-GCAT (1x50G)
750 * Mellanox(R) ConnectX(R)-4 50G MCX414A-BCAT (2x50G)
751 * Mellanox(R) ConnectX(R)-4 50G MCX415A-GCAT (2x50G)
752 * Mellanox(R) ConnectX(R)-4 50G MCX416A-BCAT (2x50G)
753 * Mellanox(R) ConnectX(R)-4 50G MCX416A-GCAT (2x50G)
754 * Mellanox(R) ConnectX(R)-4 50G MCX415A-CCAT (1x100G)
755 * Mellanox(R) ConnectX(R)-4 100G MCX416A-CCAT (2x100G)
756 * Mellanox(R) ConnectX(R)-4 Lx 10G MCX4121A-XCAT (2x10G)
757 * Mellanox(R) ConnectX(R)-4 Lx 25G MCX4121A-ACAT (2x25G)
758 * Mellanox(R) ConnectX(R)-5 100G MCX556A-ECAT (2x100G)
759 * Mellanox(R) ConnectX(R)-5 Ex EN 100G MCX516A-CDAT (2x100G)
761 Quick Start Guide on OFED/EN
762 ----------------------------
764 1. Download latest Mellanox OFED/EN. For more info check the `prerequisites`_.
767 2. Install the required libraries and kernel modules either by installing
768 only the required set, or by installing the entire Mellanox OFED/EN:
770 .. code-block:: console
772 ./mlnxofedinstall --upstream-libs --dpdk
774 3. Verify the firmware is the correct one:
776 .. code-block:: console
780 4. Verify all ports links are set to Ethernet:
782 .. code-block:: console
784 mlxconfig -d <mst device> query | grep LINK_TYPE
788 Link types may have to be configured to Ethernet:
790 .. code-block:: console
792 mlxconfig -d <mst device> set LINK_TYPE_P1/2=1/2/3
794 * LINK_TYPE_P1=<1|2|3> , 1=Infiniband 2=Ethernet 3=VPI(auto-sense)
796 For hypervisors verify SR-IOV is enabled on the NIC:
798 .. code-block:: console
800 mlxconfig -d <mst device> query | grep SRIOV_EN
803 If needed, set enable the set the relevant fields:
805 .. code-block:: console
807 mlxconfig -d <mst device> set SRIOV_EN=1 NUM_OF_VFS=16
808 mlxfwreset -d <mst device> reset
810 5. Restart the driver:
812 .. code-block:: console
814 /etc/init.d/openibd restart
818 .. code-block:: console
820 service openibd restart
822 If link type was changed, firmware must be reset as well:
824 .. code-block:: console
826 mlxfwreset -d <mst device> reset
828 For hypervisors, after reset write the sysfs number of virtual functions
831 To dynamically instantiate a given number of virtual functions (VFs):
833 .. code-block:: console
835 echo [num_vfs] > /sys/class/infiniband/mlx5_0/device/sriov_numvfs
837 6. Compile DPDK and you are ready to go. See instructions on
838 :ref:`Development Kit Build System <Development_Kit_Build_System>`
840 Enable switchdev mode
841 ---------------------
843 Switchdev mode is a mode in E-Switch, that binds between representor and VF.
844 Representor is a port in DPDK that is connected to a VF in such a way
845 that assuming there are no offload flows, each packet that is sent from the VF
846 will be received by the corresponding representor. While each packet that is
847 sent to a representor will be received by the VF.
848 This is very useful in case of SRIOV mode, where the first packet that is sent
849 by the VF will be received by the DPDK application which will decide if this
850 flow should be offloaded to the E-Switch. After offloading the flow packet
851 that the VF that are matching the flow will not be received any more by
852 the DPDK application.
854 1. Enable SRIOV mode:
856 .. code-block:: console
858 mlxconfig -d <mst device> set SRIOV_EN=true
860 2. Configure the max number of VFs:
862 .. code-block:: console
864 mlxconfig -d <mst device> set NUM_OF_VFS=<num of vfs>
868 .. code-block:: console
870 mlxfwreset -d <mst device> reset
872 3. Configure the actual number of VFs:
874 .. code-block:: console
876 echo <num of vfs > /sys/class/net/<net device>/device/sriov_numvfs
878 4. Unbind the device (can be rebind after the switchdev mode):
880 .. code-block:: console
882 echo -n "<device pci address" > /sys/bus/pci/drivers/mlx5_core/unbind
884 5. Enbale switchdev mode:
886 .. code-block:: console
888 echo switchdev > /sys/class/net/<net device>/compat/devlink/mode
893 1. Configure aggressive CQE Zipping for maximum performance:
895 .. code-block:: console
897 mlxconfig -d <mst device> s CQE_COMPRESSION=1
899 To set it back to the default CQE Zipping mode use:
901 .. code-block:: console
903 mlxconfig -d <mst device> s CQE_COMPRESSION=0
905 2. In case of virtualization:
907 - Make sure that hypervisor kernel is 3.16 or newer.
908 - Configure boot with ``iommu=pt``.
910 - Make sure to allocate a VM on huge pages.
911 - Make sure to set CPU pinning.
913 3. Use the CPU near local NUMA node to which the PCIe adapter is connected,
914 for better performance. For VMs, verify that the right CPU
915 and NUMA node are pinned according to the above. Run:
917 .. code-block:: console
921 to identify the NUMA node to which the PCIe adapter is connected.
923 4. If more than one adapter is used, and root complex capabilities allow
924 to put both adapters on the same NUMA node without PCI bandwidth degradation,
925 it is recommended to locate both adapters on the same NUMA node.
926 This in order to forward packets from one to the other without
927 NUMA performance penalty.
929 5. Disable pause frames:
931 .. code-block:: console
933 ethtool -A <netdev> rx off tx off
935 6. Verify IO non-posted prefetch is disabled by default. This can be checked
936 via the BIOS configuration. Please contact you server provider for more
937 information about the settings.
941 On some machines, depends on the machine integrator, it is beneficial
942 to set the PCI max read request parameter to 1K. This can be
943 done in the following way:
945 To query the read request size use:
947 .. code-block:: console
949 setpci -s <NIC PCI address> 68.w
951 If the output is different than 3XXX, set it by:
953 .. code-block:: console
955 setpci -s <NIC PCI address> 68.w=3XXX
957 The XXX can be different on different systems. Make sure to configure
958 according to the setpci output.
960 7. To minimize overhead of searching Memory Regions:
962 - '--socket-mem' is recommended to pin memory by predictable amount.
963 - Configure per-lcore cache when creating Mempools for packet buffer.
964 - Refrain from dynamically allocating/freeing memory in run-time.
966 Supported hardware offloads using rte_flow API
967 ----------------------------------------------
969 .. _Supported hardware offloads using rte_flow API:
971 .. table:: Supported hardware offloads using rte_flow API
973 +-----------------------+-----------------+-----------------+
974 | Offload | E-Switch | NIC |
976 +=======================+=================+=================+
977 | Count | | DPDK 19.05 | | DPDK 19.02 |
978 | | | OFED 4.6 | | OFED 4.6 |
979 | | | RDMA-CORE V24 | | RDMA-CORE V23 |
980 | | | ConnectX-5 | | ConnectX-5 |
981 +-----------------------+-----------------+-----------------+
982 | Drop / Queue / RSS | | DPDK 19.05 | | DPDK 18.11 |
983 | | | OFED 4.6 | | OFED 4.5 |
984 | | | RDMA-CORE V24 | | RDMA-CORE V23 |
985 | | | ConnectX-5 | | ConnectX-4 |
986 +-----------------------+-----------------+-----------------+
987 | Encapsulation | | DPDK 19.05 | | DPDK 19.02 |
988 | (VXLAN / NVGRE / RAW) | | OFED 4.6.2 | | OFED 4.6 |
989 | | | RDMA-CORE V24 | | RDMA-CORE V23 |
990 | | | ConnectX-5 | | ConnectX-5 |
991 +-----------------------+-----------------+-----------------+
992 | Header rewrite | | DPDK 19.05 | | DPDK 19.02 |
993 | (set_ipv4_src / | | OFED 4.6.2 | | OFED 4.6.2 |
994 | set_ipv4_dst / | | RDMA-CORE V24 | | RDMA-CORE V23 |
995 | set_ipv6_src / | | ConnectX-5 | | ConnectX-5 |
996 | set_ipv6_dst / | | |
1001 | set_mac_src / | | |
1002 | set_mac_dst) | | |
1003 +-----------------------+-----------------+-----------------+
1004 | Jump | | DPDK 19.05 | | DPDK 19.02 |
1005 | | | OFED 4.6.2 | | OFED 4.6.2 |
1006 | | | RDMA-CORE V24 | | N/A |
1007 | | | ConnectX-5 | | ConnectX-5 |
1008 +-----------------------+-----------------+-----------------+
1009 | Mark / Flag | | DPDK 19.05 | | DPDK 18.11 |
1010 | | | OFED 4.6 | | OFED 4.5 |
1011 | | | RDMA-CORE V24 | | RDMA-CORE V23 |
1012 | | | ConnectX-5 | | ConnectX-4 |
1013 +-----------------------+-----------------+-----------------+
1014 | Port ID | | DPDK 19.05 | | N/A |
1015 | | | OFED 4.6 | | N/A |
1016 | | | RDMA-CORE V24 | | N/A |
1017 | | | ConnectX-5 | | N/A |
1018 +-----------------------+-----------------+-----------------+
1020 * Minimum version for each component and nic.
1025 Compared to librte_pmd_mlx4 that implements a single RSS configuration per
1026 port, librte_pmd_mlx5 supports per-protocol RSS configuration.
1028 Since ``testpmd`` defaults to IP RSS mode and there is currently no
1029 command-line parameter to enable additional protocols (UDP and TCP as well
1030 as IP), the following commands must be entered from its CLI to get the same
1031 behavior as librte_pmd_mlx4:
1033 .. code-block:: console
1036 > port config all rss all
1042 This section demonstrates how to launch **testpmd** with Mellanox
1043 ConnectX-4/ConnectX-5/ConnectX-6/BlueField devices managed by librte_pmd_mlx5.
1045 #. Load the kernel modules:
1047 .. code-block:: console
1049 modprobe -a ib_uverbs mlx5_core mlx5_ib
1051 Alternatively if MLNX_OFED/MLNX_EN is fully installed, the following script
1054 .. code-block:: console
1056 /etc/init.d/openibd restart
1060 User space I/O kernel modules (uio and igb_uio) are not used and do
1061 not have to be loaded.
1063 #. Make sure Ethernet interfaces are in working order and linked to kernel
1064 verbs. Related sysfs entries should be present:
1066 .. code-block:: console
1068 ls -d /sys/class/net/*/device/infiniband_verbs/uverbs* | cut -d / -f 5
1072 .. code-block:: console
1079 #. Optionally, retrieve their PCI bus addresses for whitelisting:
1081 .. code-block:: console
1084 for intf in eth2 eth3 eth4 eth5;
1086 (cd "/sys/class/net/${intf}/device/" && pwd -P);
1089 sed -n 's,.*/\(.*\),-w \1,p'
1093 .. code-block:: console
1100 #. Request huge pages:
1102 .. code-block:: console
1104 echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages/nr_hugepages
1106 #. Start testpmd with basic parameters:
1108 .. code-block:: console
1110 testpmd -l 8-15 -n 4 -w 05:00.0 -w 05:00.1 -w 06:00.0 -w 06:00.1 -- --rxq=2 --txq=2 -i
1114 .. code-block:: console
1117 EAL: PCI device 0000:05:00.0 on NUMA socket 0
1118 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
1119 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_0" (VF: false)
1120 PMD: librte_pmd_mlx5: 1 port(s) detected
1121 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:fe
1122 EAL: PCI device 0000:05:00.1 on NUMA socket 0
1123 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
1124 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_1" (VF: false)
1125 PMD: librte_pmd_mlx5: 1 port(s) detected
1126 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:ff
1127 EAL: PCI device 0000:06:00.0 on NUMA socket 0
1128 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
1129 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_2" (VF: false)
1130 PMD: librte_pmd_mlx5: 1 port(s) detected
1131 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:fa
1132 EAL: PCI device 0000:06:00.1 on NUMA socket 0
1133 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
1134 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_3" (VF: false)
1135 PMD: librte_pmd_mlx5: 1 port(s) detected
1136 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:fb
1137 Interactive-mode selected
1138 Configuring Port 0 (socket 0)
1139 PMD: librte_pmd_mlx5: 0x8cba80: TX queues number update: 0 -> 2
1140 PMD: librte_pmd_mlx5: 0x8cba80: RX queues number update: 0 -> 2
1141 Port 0: E4:1D:2D:E7:0C:FE
1142 Configuring Port 1 (socket 0)
1143 PMD: librte_pmd_mlx5: 0x8ccac8: TX queues number update: 0 -> 2
1144 PMD: librte_pmd_mlx5: 0x8ccac8: RX queues number update: 0 -> 2
1145 Port 1: E4:1D:2D:E7:0C:FF
1146 Configuring Port 2 (socket 0)
1147 PMD: librte_pmd_mlx5: 0x8cdb10: TX queues number update: 0 -> 2
1148 PMD: librte_pmd_mlx5: 0x8cdb10: RX queues number update: 0 -> 2
1149 Port 2: E4:1D:2D:E7:0C:FA
1150 Configuring Port 3 (socket 0)
1151 PMD: librte_pmd_mlx5: 0x8ceb58: TX queues number update: 0 -> 2
1152 PMD: librte_pmd_mlx5: 0x8ceb58: RX queues number update: 0 -> 2
1153 Port 3: E4:1D:2D:E7:0C:FB
1154 Checking link statuses...
1155 Port 0 Link Up - speed 40000 Mbps - full-duplex
1156 Port 1 Link Up - speed 40000 Mbps - full-duplex
1157 Port 2 Link Up - speed 10000 Mbps - full-duplex
1158 Port 3 Link Up - speed 10000 Mbps - full-duplex