-.. BSD LICENSE
- Copyright 2012-2015 6WIND S.A.
-
- 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 6WIND S.A. 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.
+.. SPDX-License-Identifier: BSD-3-Clause
+ Copyright 2012 6WIND S.A.
+ Copyright 2015 Mellanox Technologies, Ltd
MLX4 poll mode driver library
=============================
The MLX4 poll mode driver library (**librte_pmd_mlx4**) implements support
-for **Mellanox ConnectX-3** 10/40 Gbps adapters (EN 40, EN 10, Pro EN 40) as
-well as their virtual functions (VF) in SR-IOV context.
+for **Mellanox ConnectX-3** and **Mellanox ConnectX-3 Pro** 10/40 Gbps adapters
+as well as their virtual functions (VF) in SR-IOV context.
+
+Information and documentation about this family of adapters can be found on
+the `Mellanox website <http://www.mellanox.com>`_. Help is also provided by
+the `Mellanox community <http://community.mellanox.com/welcome>`_.
+
+There is also a `section dedicated to this poll mode driver
+<http://www.mellanox.com/page/products_dyn?product_family=209&mtag=pmd_for_dpdk>`_.
.. note::
which remain functional, although they stop receiving unicast packets as
long as they share the same MAC address.
-Compiling librte_pmd_mlx4 causes DPDK to be linked against libibverbs.
-
-Features and limitations
-------------------------
+The :ref:`flow_isolated_mode` is supported.
-- RSS, also known as RCA, is supported. In this mode the number of
- configured RX queues must be a power of two.
-- VLAN filtering is supported.
-- Link state information is provided.
-- Promiscuous mode is supported.
-- All multicast mode is supported.
-- Multiple MAC addresses (unicast, multicast) can be configured.
-- Scattered packets are supported for TX and RX.
-
-..
-
-- RSS hash key cannot be modified.
-- Hardware counters are not implemented (they are software counters).
-- Checksum offloads are not supported yet.
+Compiling librte_pmd_mlx4 causes DPDK to be linked against libibverbs.
Configuration
-------------
Compilation options
~~~~~~~~~~~~~~~~~~~
+These options can be modified in the ``.config`` file.
+
- ``CONFIG_RTE_LIBRTE_MLX4_PMD`` (default **n**)
Toggle compilation of librte_pmd_mlx4 itself.
-- ``CONFIG_RTE_LIBRTE_MLX4_DEBUG`` (default **n**)
+- ``CONFIG_RTE_IBVERBS_LINK_DLOPEN`` (default **n**)
- Toggle debugging code and stricter compilation flags. Enabling this option
- adds additional run-time checks and debugging messages at the cost of
- lower performance.
+ Build PMD with additional code to make it loadable without hard
+ dependencies on **libibverbs** nor **libmlx4**, which may not be installed
+ on the target system.
-- ``CONFIG_RTE_LIBRTE_MLX4_SGE_WR_N`` (default **4**)
+ In this mode, their presence is still required for it to run properly,
+ however their absence won't prevent a DPDK application from starting (with
+ ``CONFIG_RTE_BUILD_SHARED_LIB`` disabled) and they won't show up as
+ missing with ``ldd(1)``.
- Number of scatter/gather elements (SGEs) per work request (WR). Lowering
- this number improves performance but also limits the ability to receive
- scattered packets (packets that do not fit a single mbuf). The default
- value is a safe tradeoff.
+ It works by moving these dependencies to a purpose-built rdma-core "glue"
+ plug-in which must either be installed in a directory whose name is based
+ on ``CONFIG_RTE_EAL_PMD_PATH`` suffixed with ``-glue`` if set, or in a
+ standard location for the dynamic linker (e.g. ``/lib``) if left to the
+ default empty string (``""``).
-- ``CONFIG_RTE_LIBRTE_MLX4_MAX_INLINE`` (default **0**)
+ This option has no performance impact.
- Amount of data to be inlined during TX operations. Improves latency but
- lowers throughput.
+- ``CONFIG_RTE_IBVERBS_LINK_STATIC`` (default **n**)
-- ``CONFIG_RTE_LIBRTE_MLX4_TX_MP_CACHE`` (default **8**)
+ Embed static flavor of the dependencies **libibverbs** and **libmlx4**
+ in the PMD shared library or the executable static binary.
- Maximum number of cached memory pools (MPs) per TX queue. Each MP from
- which buffers are to be transmitted must be associated to memory regions
- (MRs). This is a slow operation that must be cached.
+- ``CONFIG_RTE_LIBRTE_MLX4_DEBUG`` (default **n**)
- This value is always 1 for RX queues since they use a single MP.
+ Toggle debugging code and stricter compilation flags. Enabling this option
+ adds additional run-time checks and debugging messages at the cost of
+ lower performance.
-- ``CONFIG_RTE_LIBRTE_MLX4_SOFT_COUNTERS`` (default **1**)
+This option is available in meson:
- Toggle software counters. No counters are available if this option is
- disabled since hardware counters are not supported.
+- ``ibverbs_link`` can be ``static``, ``shared``, or ``dlopen``.
Environment variables
~~~~~~~~~~~~~~~~~~~~~
-- ``MLX4_INLINE_RECV_SIZE``
+- ``MLX4_GLUE_PATH``
+
+ A list of directories in which to search for the rdma-core "glue" plug-in,
+ separated by colons or semi-colons.
- A nonzero value enables inline receive for packets up to that size. May
- significantly improve performance in some cases but lower it in
- others. Requires careful testing.
+ Only matters when compiled with ``CONFIG_RTE_IBVERBS_LINK_DLOPEN``
+ enabled and most useful when ``CONFIG_RTE_EAL_PMD_PATH`` is also set,
+ since ``LD_LIBRARY_PATH`` has no effect in this case.
Run-time configuration
~~~~~~~~~~~~~~~~~~~~~~
-- The only constraint when RSS mode is requested is to make sure the number
- of RX queues is a power of two. This is a hardware requirement.
-
- librte_pmd_mlx4 brings kernel network interfaces up during initialization
because it is affected by their state. Forcing them down prevents packets
reception.
- **ethtool** operations on related kernel interfaces also affect the PMD.
+- ``port`` parameter [int]
+
+ This parameter provides a physical port to probe and can be specified multiple
+ times for additional ports. All ports are probed by default if left
+ unspecified.
+
+- ``mr_ext_memseg_en`` parameter [int]
+
+ A nonzero value enables extending memseg when registering DMA memory. If
+ enabled, the number of entries in MR (Memory Region) lookup table on datapath
+ is minimized and it benefits performance. On the other hand, it worsens memory
+ utilization because registered memory is pinned by kernel driver. Even if a
+ page in the extended chunk is freed, that doesn't become reusable until the
+ entire memory is freed.
+
+ Enabled by default.
+
+Kernel module parameters
+~~~~~~~~~~~~~~~~~~~~~~~~
+
+The **mlx4_core** kernel module has several parameters that affect the
+behavior and/or the performance of librte_pmd_mlx4. Some of them are described
+below.
+
+- **num_vfs** (integer or triplet, optionally prefixed by device address
+ strings)
+
+ Create the given number of VFs on the specified devices.
+
+- **log_num_mgm_entry_size** (integer)
+
+ Device-managed flow steering (DMFS) is required by DPDK applications. It is
+ enabled by using a negative value, the last four bits of which have a
+ special meaning.
+
+ - **-1**: force device-managed flow steering (DMFS).
+ - **-7**: configure optimized steering mode to improve performance with the
+ following limitation: VLAN filtering is not supported with this mode.
+ This is the recommended mode in case VLAN filter is not needed.
+
+Limitations
+-----------
+
+- For secondary process:
+
+ - Forked secondary process not supported.
+ - External memory unregistered in EAL memseg list cannot be used for DMA
+ unless such memory has been registered by ``mlx4_mr_update_ext_mp()`` in
+ primary process and remapped to the same virtual address in secondary
+ process. If the external memory is registered by primary process but has
+ different virtual address in secondary process, unexpected error may happen.
+
+- CRC stripping is supported by default and always reported as "true".
+ The ability to enable/disable CRC stripping requires OFED version
+ 4.3-1.5.0.0 and above or rdma-core version v18 and above.
+
+- TSO (Transmit Segmentation Offload) is supported in OFED version
+ 4.4 and above.
+
Prerequisites
-------------
allocations and initialization. The following dependencies are not part of
DPDK and must be installed separately:
-- **libibverbs**
+- **libibverbs** (provided by rdma-core package)
User space verbs framework used by librte_pmd_mlx4. This library provides
a generic interface between the kernel and low-level user space drivers
resources allocations) to be managed by the kernel and fast operations to
never leave user space.
-- **libmlx4**
+- **libmlx4** (provided by rdma-core package)
Low-level user space driver library for Mellanox ConnectX-3 devices,
it is automatically loaded by libibverbs.
This library basically implements send/receive calls to the hardware
queues.
-- **Kernel modules** (mlnx-ofed-kernel)
+- **Kernel modules**
They provide the kernel-side verbs API and low level device drivers that
manage actual hardware initialization and resources sharing with user
- mlx4_ib: InifiniBand device driver.
- ib_uverbs: user space driver for verbs (entry point for libibverbs).
-While these libraries and kernel modules are available on OpenFabrics
-Aliance's `website <https://www.openfabrics.org/>`_ and provided by package
-managers on most distributions, this PMD requires Ethernet extensions that
-may not be supported at the moment (this is a work in progress).
+- **Firmware update**
+
+ Mellanox OFED releases include firmware updates for ConnectX-3 adapters.
-`Mellanox OFED
-<http://www.mellanox.com/page/products_dyn?product_family=26&mtag=linux_sw_drivers>`_
-includes the necessary support and should be used in the meantime. For DPDK,
-only libibverbs, libmlx4 and mlnx-ofed-kernel packages are required from
-that distribution.
+ Because each release provides new features, these updates must be applied to
+ match the kernel modules and libraries they come with.
.. note::
Both libraries are BSD and GPL licensed. Linux kernel modules are GPL
licensed.
+Depending on system constraints and user preferences either RDMA core library
+with a recent enough Linux kernel release (recommended) or Mellanox OFED,
+which provides compatibility with older releases.
+
+Current RDMA core package and Linux kernel (recommended)
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+- Minimal Linux kernel version: 4.14.
+- Minimal RDMA core version: v15 (see `RDMA core installation documentation`_).
+
+- Starting with rdma-core v21, static libraries can be built::
+
+ cd build
+ CFLAGS=-fPIC cmake -DIN_PLACE=1 -DENABLE_STATIC=1 -GNinja ..
+ ninja
+
+.. _`RDMA core installation documentation`: https://raw.githubusercontent.com/linux-rdma/rdma-core/master/README.md
+
+If rdma-core libraries are built but not installed, DPDK makefile can link them,
+thanks to these environment variables:
+
+ - ``EXTRA_CFLAGS=-I/path/to/rdma-core/build/include``
+ - ``EXTRA_LDFLAGS=-L/path/to/rdma-core/build/lib``
+ - ``PKG_CONFIG_PATH=/path/to/rdma-core/build/lib/pkgconfig``
+
+.. _Mellanox_OFED_as_a_fallback:
+
+Mellanox OFED as a fallback
+~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+- `Mellanox OFED`_ version: **4.4, 4.5, 4.6**.
+- firmware version: **2.42.5000** and above.
+
+.. _`Mellanox OFED`: http://www.mellanox.com/page/products_dyn?product_family=26&mtag=linux_sw_drivers
+
+.. note::
+
+ Several versions of Mellanox OFED are available. Installing the version
+ this DPDK release was developed and tested against is strongly
+ recommended. Please check the `prerequisites`_.
+
+Installing Mellanox OFED
+^^^^^^^^^^^^^^^^^^^^^^^^
+
+1. Download latest Mellanox OFED.
+
+2. Install the required libraries and kernel modules either by installing
+ only the required set, or by installing the entire Mellanox OFED:
+
+ For bare metal use::
+
+ ./mlnxofedinstall --dpdk --upstream-libs
+
+ For SR-IOV hypervisors use::
+
+ ./mlnxofedinstall --dpdk --upstream-libs --enable-sriov --hypervisor
+
+ For SR-IOV virtual machine use::
+
+ ./mlnxofedinstall --dpdk --upstream-libs --guest
+
+3. Verify the firmware is the correct one::
+
+ ibv_devinfo
+
+4. Set all ports links to Ethernet, follow instructions on the screen::
+
+ connectx_port_config
+
+5. Continue with :ref:`section 2 of the Quick Start Guide <QSG_2>`.
+
+.. _qsg:
+
+Quick Start Guide
+-----------------
+
+1. Set all ports links to Ethernet::
+
+ PCI=<NIC PCI address>
+ echo eth > "/sys/bus/pci/devices/$PCI/mlx4_port0"
+ echo eth > "/sys/bus/pci/devices/$PCI/mlx4_port1"
+
+ .. note::
+
+ If using Mellanox OFED one can permanently set the port link
+ to Ethernet using connectx_port_config tool provided by it.
+ :ref:`Mellanox_OFED_as_a_fallback`:
+
+.. _QSG_2:
+
+2. In case of bare metal or hypervisor, configure optimized steering mode
+ by adding the following line to ``/etc/modprobe.d/mlx4_core.conf``::
+
+ options mlx4_core log_num_mgm_entry_size=-7
+
+ .. note::
+
+ If VLAN filtering is used, set log_num_mgm_entry_size=-1.
+ Performance degradation can occur on this case.
+
+3. Restart the driver::
+
+ /etc/init.d/openibd restart
+
+ or::
+
+ service openibd restart
+
+4. Compile DPDK and you are ready to go. See instructions on
+ :ref:`Development Kit Build System <Development_Kit_Build_System>`
+
+Performance tuning
+------------------
+
+1. Verify the optimized steering mode is configured::
+
+ cat /sys/module/mlx4_core/parameters/log_num_mgm_entry_size
+
+2. Use the CPU near local NUMA node to which the PCIe adapter is connected,
+ for better performance. For VMs, verify that the right CPU
+ and NUMA node are pinned according to the above. Run::
+
+ lstopo-no-graphics
+
+ to identify the NUMA node to which the PCIe adapter is connected.
+
+3. If more than one adapter is used, and root complex capabilities allow
+ to put both adapters on the same NUMA node without PCI bandwidth degradation,
+ it is recommended to locate both adapters on the same NUMA node.
+ This in order to forward packets from one to the other without
+ NUMA performance penalty.
+
+4. Disable pause frames::
+
+ ethtool -A <netdev> rx off tx off
+
+5. Verify IO non-posted prefetch is disabled by default. This can be checked
+ via the BIOS configuration. Please contact you server provider for more
+ information about the settings.
+
+.. note::
+
+ On some machines, depends on the machine integrator, it is beneficial
+ to set the PCI max read request parameter to 1K. This can be
+ done in the following way:
+
+ To query the read request size use::
+
+ setpci -s <NIC PCI address> 68.w
+
+ If the output is different than 3XXX, set it by::
+
+ setpci -s <NIC PCI address> 68.w=3XXX
+
+ The XXX can be different on different systems. Make sure to configure
+ according to the setpci output.
+
+6. To minimize overhead of searching Memory Regions:
+
+ - '--socket-mem' is recommended to pin memory by predictable amount.
+ - Configure per-lcore cache when creating Mempools for packet buffer.
+ - Refrain from dynamically allocating/freeing memory in run-time.
+
Usage example
-------------
This section demonstrates how to launch **testpmd** with Mellanox ConnectX-3
devices managed by librte_pmd_mlx4.
-#. Load the kernel modules:
-
- .. code-block:: console
+#. Load the kernel modules::
modprobe -a ib_uverbs mlx4_en mlx4_core mlx4_ib
+ Alternatively if MLNX_OFED is fully installed, the following script can
+ be run::
+
+ /etc/init.d/openibd restart
+
.. note::
User space I/O kernel modules (uio and igb_uio) are not used and do
not have to be loaded.
#. Make sure Ethernet interfaces are in working order and linked to kernel
- verbs. Related sysfs entries should be present:
-
- .. code-block:: console
+ verbs. Related sysfs entries should be present::
ls -d /sys/class/net/*/device/infiniband_verbs/uverbs* | cut -d / -f 5
- Example output:
-
- .. code-block:: console
+ Example output::
eth2
eth3
eth4
eth5
-#. Optionally, retrieve their PCI bus addresses for whitelisting:
-
- .. code-block:: console
+#. Optionally, retrieve their PCI bus addresses for whitelisting::
{
for intf in eth2 eth3 eth4 eth5;
} |
sed -n 's,.*/\(.*\),-w \1,p'
- Example output:
-
- .. code-block:: console
+ Example output::
-w 0000:83:00.0
-w 0000:83:00.0
There are only two distinct PCI bus addresses because the Mellanox
ConnectX-3 adapters installed on this system are dual port.
-#. Request huge pages:
-
- .. code-block:: console
+#. Request huge pages::
echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages/nr_hugepages
-#. Start testpmd with basic parameters:
-
- .. code-block:: console
-
- testpmd -c 0xff00 -n 4 -w 0000:83:00.0 -w 0000:84:00.0 -- --rxq=2 --txq=2 -i
+#. Start testpmd with basic parameters::
- Example output:
+ testpmd -l 8-15 -n 4 -w 0000:83:00.0 -w 0000:84:00.0 -- --rxq=2 --txq=2 -i
- .. code-block:: console
+ Example output::
[...]
EAL: PCI device 0000:83:00.0 on NUMA socket 1