2 Copyright 2015 6WIND S.A.
4 Redistribution and use in source and binary forms, with or without
5 modification, are permitted provided that the following conditions
8 * Redistributions of source code must retain the above copyright
9 notice, this list of conditions and the following disclaimer.
10 * Redistributions in binary form must reproduce the above copyright
11 notice, this list of conditions and the following disclaimer in
12 the documentation and/or other materials provided with the
14 * Neither the name of 6WIND S.A. nor the names of its
15 contributors may be used to endorse or promote products derived
16 from this software without specific prior written permission.
18 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
19 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
20 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
21 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
22 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
23 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
24 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
25 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
26 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
27 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
28 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 The MLX5 poll mode driver library (**librte_pmd_mlx5**) provides support for
34 **Mellanox ConnectX-4** and **Mellanox ConnectX-4 Lx** families of
35 10/25/40/50/100 Gb/s adapters as well as their virtual functions (VF) in
38 Information and documentation about these adapters can be found on the
39 `Mellanox website <http://www.mellanox.com>`__. Help is also provided by the
40 `Mellanox community <http://community.mellanox.com/welcome>`__.
42 There is also a `section dedicated to this poll mode driver
43 <http://www.mellanox.com/page/products_dyn?product_family=209&mtag=pmd_for_dpdk>`__.
47 Due to external dependencies, this driver is disabled by default. It must
48 be enabled manually by setting ``CONFIG_RTE_LIBRTE_MLX5_PMD=y`` and
51 Implementation details
52 ----------------------
54 Besides its dependency on libibverbs (that implies libmlx5 and associated
55 kernel support), librte_pmd_mlx5 relies heavily on system calls for control
56 operations such as querying/updating the MTU and flow control parameters.
58 For security reasons and robustness, this driver only deals with virtual
59 memory addresses. The way resources allocations are handled by the kernel
60 combined with hardware specifications that allow it to handle virtual memory
61 addresses directly ensure that DPDK applications cannot access random
62 physical memory (or memory that does not belong to the current process).
64 This capability allows the PMD to coexist with kernel network interfaces
65 which remain functional, although they stop receiving unicast packets as
66 long as they share the same MAC address.
68 Enabling librte_pmd_mlx5 causes DPDK applications to be linked against
74 - Multiple TX and RX queues.
75 - Support for scattered TX and RX frames.
76 - IPv4, IPv6, TCPv4, TCPv6, UDPv4 and UDPv6 RSS on any number of queues.
77 - Several RSS hash keys, one for each flow type.
78 - Configurable RETA table.
79 - Support for multiple MAC addresses.
83 - RX CRC stripping configuration.
85 - Multicast promiscuous mode.
86 - Hardware checksum offloads.
87 - Flow director (RTE_FDIR_MODE_PERFECT and RTE_FDIR_MODE_PERFECT_MAC_VLAN).
88 - Secondary process TX is supported.
89 - KVM and VMware ESX SR-IOV modes are supported.
94 - Inner RSS for VXLAN frames is not supported yet.
95 - Port statistics through software counters only.
96 - Hardware checksum offloads for VXLAN inner header are not supported yet.
97 - Secondary process RX is not supported.
105 These options can be modified in the ``.config`` file.
107 - ``CONFIG_RTE_LIBRTE_MLX5_PMD`` (default **n**)
109 Toggle compilation of librte_pmd_mlx5 itself.
111 - ``CONFIG_RTE_LIBRTE_MLX5_DEBUG`` (default **n**)
113 Toggle debugging code and stricter compilation flags. Enabling this option
114 adds additional run-time checks and debugging messages at the cost of
117 - ``CONFIG_RTE_LIBRTE_MLX5_TX_MP_CACHE`` (default **8**)
119 Maximum number of cached memory pools (MPs) per TX queue. Each MP from
120 which buffers are to be transmitted must be associated to memory regions
121 (MRs). This is a slow operation that must be cached.
123 This value is always 1 for RX queues since they use a single MP.
125 Environment variables
126 ~~~~~~~~~~~~~~~~~~~~~
128 - ``MLX5_PMD_ENABLE_PADDING``
130 Enables HW packet padding in PCI bus transactions.
132 When packet size is cache aligned and CRC stripping is enabled, 4 fewer
133 bytes are written to the PCI bus. Enabling padding makes such packets
136 In cases where PCI bandwidth is the bottleneck, padding can improve
139 This is disabled by default since this can also decrease performance for
140 unaligned packet sizes.
142 Run-time configuration
143 ~~~~~~~~~~~~~~~~~~~~~~
145 - librte_pmd_mlx5 brings kernel network interfaces up during initialization
146 because it is affected by their state. Forcing them down prevents packets
149 - **ethtool** operations on related kernel interfaces also affect the PMD.
151 - ``rxq_cqe_comp_en`` parameter [int]
153 A nonzero value enables the compression of CQE on RX side. This feature
154 allows to save PCI bandwidth and improve performance at the cost of a
155 slightly higher CPU usage. Enabled by default.
157 - ``txq_inline`` parameter [int]
159 Amount of data to be inlined during TX operations. Improves latency.
160 Can improve PPS performance when PCI back pressure is detected and may be
161 useful for scenarios involving heavy traffic on many queues.
163 It is not enabled by default (set to 0) since the additional software
164 logic necessary to handle this mode can lower performance when back
165 pressure is not expected.
167 - ``txqs_min_inline`` parameter [int]
169 Enable inline send only when the number of TX queues is greater or equal
172 This option should be used in combination with ``txq_inline`` above.
177 This driver relies on external libraries and kernel drivers for resources
178 allocations and initialization. The following dependencies are not part of
179 DPDK and must be installed separately:
183 User space Verbs framework used by librte_pmd_mlx5. This library provides
184 a generic interface between the kernel and low-level user space drivers
187 It allows slow and privileged operations (context initialization, hardware
188 resources allocations) to be managed by the kernel and fast operations to
189 never leave user space.
193 Low-level user space driver library for Mellanox ConnectX-4 devices,
194 it is automatically loaded by libibverbs.
196 This library basically implements send/receive calls to the hardware
199 - **Kernel modules** (mlnx-ofed-kernel)
201 They provide the kernel-side Verbs API and low level device drivers that
202 manage actual hardware initialization and resources sharing with user
205 Unlike most other PMDs, these modules must remain loaded and bound to
208 - mlx5_core: hardware driver managing Mellanox ConnectX-4 devices and
209 related Ethernet kernel network devices.
210 - mlx5_ib: InifiniBand device driver.
211 - ib_uverbs: user space driver for Verbs (entry point for libibverbs).
213 - **Firmware update**
215 Mellanox OFED releases include firmware updates for ConnectX-4 adapters.
217 Because each release provides new features, these updates must be applied to
218 match the kernel modules and libraries they come with.
222 Both libraries are BSD and GPL licensed. Linux kernel modules are GPL
225 Currently supported by DPDK:
227 - Mellanox OFED **3.3-1.0.0.0**.
229 - Minimum firmware version:
231 - ConnectX-4: **12.16.1006**
232 - ConnectX-4 Lx: **14.16.1006**
234 Getting Mellanox OFED
235 ~~~~~~~~~~~~~~~~~~~~~
237 While these libraries and kernel modules are available on OpenFabrics
238 Alliance's `website <https://www.openfabrics.org/>`__ and provided by package
239 managers on most distributions, this PMD requires Ethernet extensions that
240 may not be supported at the moment (this is a work in progress).
243 <http://www.mellanox.com/page/products_dyn?product_family=26&mtag=linux>`__
244 includes the necessary support and should be used in the meantime. For DPDK,
245 only libibverbs, libmlx5, mlnx-ofed-kernel packages and firmware updates are
246 required from that distribution.
250 Several versions of Mellanox OFED are available. Installing the version
251 this DPDK release was developed and tested against is strongly
252 recommended. Please check the `prerequisites`_.
257 Compared to librte_pmd_mlx4 that implements a single RSS configuration per
258 port, librte_pmd_mlx5 supports per-protocol RSS configuration.
260 Since ``testpmd`` defaults to IP RSS mode and there is currently no
261 command-line parameter to enable additional protocols (UDP and TCP as well
262 as IP), the following commands must be entered from its CLI to get the same
263 behavior as librte_pmd_mlx4:
265 .. code-block:: console
268 > port config all rss all
274 This section demonstrates how to launch **testpmd** with Mellanox ConnectX-4
275 devices managed by librte_pmd_mlx5.
277 #. Load the kernel modules:
279 .. code-block:: console
281 modprobe -a ib_uverbs mlx5_core mlx5_ib
283 Alternatively if MLNX_OFED is fully installed, the following script can
286 .. code-block:: console
288 /etc/init.d/openibd restart
292 User space I/O kernel modules (uio and igb_uio) are not used and do
293 not have to be loaded.
295 #. Make sure Ethernet interfaces are in working order and linked to kernel
296 verbs. Related sysfs entries should be present:
298 .. code-block:: console
300 ls -d /sys/class/net/*/device/infiniband_verbs/uverbs* | cut -d / -f 5
304 .. code-block:: console
311 #. Optionally, retrieve their PCI bus addresses for whitelisting:
313 .. code-block:: console
316 for intf in eth2 eth3 eth4 eth5;
318 (cd "/sys/class/net/${intf}/device/" && pwd -P);
321 sed -n 's,.*/\(.*\),-w \1,p'
325 .. code-block:: console
332 #. Request huge pages:
334 .. code-block:: console
336 echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages/nr_hugepages
338 #. Start testpmd with basic parameters:
340 .. code-block:: console
342 testpmd -c 0xff00 -n 4 -w 05:00.0 -w 05:00.1 -w 06:00.0 -w 06:00.1 -- --rxq=2 --txq=2 -i
346 .. code-block:: console
349 EAL: PCI device 0000:05:00.0 on NUMA socket 0
350 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
351 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_0" (VF: false)
352 PMD: librte_pmd_mlx5: 1 port(s) detected
353 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:fe
354 EAL: PCI device 0000:05:00.1 on NUMA socket 0
355 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
356 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_1" (VF: false)
357 PMD: librte_pmd_mlx5: 1 port(s) detected
358 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:ff
359 EAL: PCI device 0000:06:00.0 on NUMA socket 0
360 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
361 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_2" (VF: false)
362 PMD: librte_pmd_mlx5: 1 port(s) detected
363 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:fa
364 EAL: PCI device 0000:06:00.1 on NUMA socket 0
365 EAL: probe driver: 15b3:1013 librte_pmd_mlx5
366 PMD: librte_pmd_mlx5: PCI information matches, using device "mlx5_3" (VF: false)
367 PMD: librte_pmd_mlx5: 1 port(s) detected
368 PMD: librte_pmd_mlx5: port 1 MAC address is e4:1d:2d:e7:0c:fb
369 Interactive-mode selected
370 Configuring Port 0 (socket 0)
371 PMD: librte_pmd_mlx5: 0x8cba80: TX queues number update: 0 -> 2
372 PMD: librte_pmd_mlx5: 0x8cba80: RX queues number update: 0 -> 2
373 Port 0: E4:1D:2D:E7:0C:FE
374 Configuring Port 1 (socket 0)
375 PMD: librte_pmd_mlx5: 0x8ccac8: TX queues number update: 0 -> 2
376 PMD: librte_pmd_mlx5: 0x8ccac8: RX queues number update: 0 -> 2
377 Port 1: E4:1D:2D:E7:0C:FF
378 Configuring Port 2 (socket 0)
379 PMD: librte_pmd_mlx5: 0x8cdb10: TX queues number update: 0 -> 2
380 PMD: librte_pmd_mlx5: 0x8cdb10: RX queues number update: 0 -> 2
381 Port 2: E4:1D:2D:E7:0C:FA
382 Configuring Port 3 (socket 0)
383 PMD: librte_pmd_mlx5: 0x8ceb58: TX queues number update: 0 -> 2
384 PMD: librte_pmd_mlx5: 0x8ceb58: RX queues number update: 0 -> 2
385 Port 3: E4:1D:2D:E7:0C:FB
386 Checking link statuses...
387 Port 0 Link Up - speed 40000 Mbps - full-duplex
388 Port 1 Link Up - speed 40000 Mbps - full-duplex
389 Port 2 Link Up - speed 10000 Mbps - full-duplex
390 Port 3 Link Up - speed 10000 Mbps - full-duplex