1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2015 Intel Corporation.
3 Copyright 2017 Mellanox Technologies, Ltd
6 .. include:: <isonum.txt>
8 .. _linux_gsg_linux_drivers:
13 Different PMDs may require different kernel drivers in order to work properly.
14 Depending on the PMD being used, a corresponding kernel driver should be loaded,
15 and network ports should be bound to that driver.
17 .. _linux_gsg_binding_kernel:
19 Binding and Unbinding Network Ports to/from the Kernel Modules
20 --------------------------------------------------------------
24 PMDs which use the bifurcated driver should not be unbound from their kernel drivers.
25 This section is for PMDs which use the UIO or VFIO drivers.
26 See :ref:`bifurcated_driver` section for more details.
30 It is recommended that ``vfio-pci`` be used as the kernel module for DPDK-bound ports in all cases.
31 If an IOMMU is unavailable, the ``vfio-pci`` can be used in :ref:`no-iommu<vfio_noiommu>` mode.
32 If, for some reason, vfio is unavailable, then UIO-based modules, ``igb_uio`` and ``uio_pci_generic`` may be used.
33 See section :ref:`uio` for details.
35 Most devices require that the hardware to be used by DPDK be unbound from the kernel driver it uses,
36 and instead be bound to the ``vfio-pci`` kernel module before the application is run.
37 For such PMDs, any network ports or other hardware under Linux* control will be ignored and cannot be used by the application.
39 To bind ports to the ``vfio-pci`` module
40 for DPDK use, or to return ports to Linux control,
41 a utility script called ``dpdk-devbind.py`` is provided in the ``usertools`` subdirectory.
42 This utility can be used to provide a view of the current state of the network ports on the system,
43 and to bind and unbind those ports from the different kernel modules,
44 including the VFIO and UIO modules.
45 The following are some examples of how the script can be used.
46 A full description of the script and its parameters can be obtained
47 by calling the script with the ``--help`` or ``--usage`` options.
48 Note that the UIO or VFIO kernel modules to be used,
49 should be loaded into the kernel before running the ``dpdk-devbind.py`` script.
53 Due to the way VFIO works, there are certain limitations
54 to which devices can be used with VFIO.
55 Mainly it comes down to how IOMMU groups work.
56 Any Virtual Function device can usually be used with VFIO on its own,
57 but physical devices may require either all ports bound to VFIO,
58 or some of them bound to VFIO while others not being bound to anything at all.
60 If your device is behind a PCI-to-PCI bridge,
61 the bridge will then be part of the IOMMU group in which your device is in.
62 Therefore, the bridge driver should also be unbound from the bridge PCI device
63 for VFIO to work with devices behind the bridge.
67 While any user can run the ``dpdk-devbind.py`` script
68 to view the status of the network ports,
69 binding or unbinding network ports requires root privileges.
71 To see the status of all network ports on the system:
73 .. code-block:: console
75 ./usertools/dpdk-devbind.py --status
77 Network devices using DPDK-compatible driver
78 ============================================
79 0000:82:00.0 '82599EB 10-GbE NIC' drv=vfio-pci unused=ixgbe
80 0000:82:00.1 '82599EB 10-GbE NIC' drv=vfio-pci unused=ixgbe
82 Network devices using kernel driver
83 ===================================
84 0000:04:00.0 'I350 1-GbE NIC' if=em0 drv=igb unused=vfio-pci *Active*
85 0000:04:00.1 'I350 1-GbE NIC' if=eth1 drv=igb unused=vfio-pci
86 0000:04:00.2 'I350 1-GbE NIC' if=eth2 drv=igb unused=vfio-pci
87 0000:04:00.3 'I350 1-GbE NIC' if=eth3 drv=igb unused=vfio-pci
93 To bind device ``eth1``,``04:00.1``, to the ``vfio-pci`` driver:
95 .. code-block:: console
97 ./usertools/dpdk-devbind.py --bind=vfio-pci 04:00.1
101 .. code-block:: console
103 ./usertools/dpdk-devbind.py --bind=vfio-pci eth1
105 When specifying device ids, wildcards can be used for the final part of the address.
106 To restore device ``82:00.0`` and ``82:00.1`` to their original kernel binding:
108 .. code-block:: console
110 ./usertools/dpdk-devbind.py --bind=ixgbe 82:00.*
115 VFIO is a robust and secure driver that relies on IOMMU protection.
116 To make use of VFIO, the ``vfio-pci`` module must be loaded:
118 .. code-block:: console
120 sudo modprobe vfio-pci
122 VFIO kernel is usually present by default in all distributions,
123 however please consult your distributions documentation to make sure that is the case.
125 For DMA mapping of either external memory or hugepages, VFIO interface is used.
126 VFIO does not support partial unmap of once mapped memory. Hence DPDK's memory is
127 mapped in hugepage granularity or system page granularity. Number of DMA
128 mappings is limited by kernel with user locked memory limit of a process (rlimit)
129 for system/hugepage memory. Another per-container overall limit applicable both
130 for external memory and system memory was added in kernel 5.1 defined by
131 VFIO module parameter ``dma_entry_limit`` with a default value of 64K.
132 When application is out of DMA entries, these limits need to be adjusted to
133 increase the allowed limit.
135 Since Linux version 5.7,
136 the ``vfio-pci`` module supports the creation of virtual functions.
137 After the PF is bound to ``vfio-pci`` module,
138 the user can create the VFs using the ``sysfs`` interface,
139 and these VFs will be bound to ``vfio-pci`` module automatically.
141 When the PF is bound to ``vfio-pci``,
142 by default it will have a randomly generated VF token.
143 For security reasons, this token is write only,
144 so the user cannot read it from the kernel directly.
145 To access the VFs, the user needs to create a new token,
146 and use it to initialize both VF and PF devices.
147 The tokens are in UUID format,
148 so any UUID generation tool can be used to create a new token.
150 This VF token can be passed to DPDK by using EAL parameter ``--vfio-vf-token``.
151 The token will be used for all PF and VF ports within the application.
153 #. Generate the VF token by uuid command
155 .. code-block:: console
157 14d63f20-8445-11ea-8900-1f9ce7d5650d
159 #. Load the ``vfio-pci`` module with ``enable_sriov`` parameter set
161 .. code-block:: console
163 sudo modprobe vfio-pci enable_sriov=1
165 Alternatively, pass the ``enable_sriov`` parameter through the ``sysfs`` if the module is already loaded or is built-in:
167 .. code-block:: console
169 echo 1 | sudo tee /sys/module/vfio_pci/parameters/enable_sriov
171 #. Bind the PCI devices to ``vfio-pci`` driver
173 .. code-block:: console
175 ./usertools/dpdk-devbind.py -b vfio-pci 0000:86:00.0
177 #. Create the desired number of VF devices
179 .. code-block:: console
181 echo 2 > /sys/bus/pci/devices/0000:86:00.0/sriov_numvfs
183 #. Start the DPDK application that will manage the PF device
185 .. code-block:: console
187 <build_dir>/app/dpdk-testpmd -l 22-25 -n 4 -a 86:00.0 \
188 --vfio-vf-token=14d63f20-8445-11ea-8900-1f9ce7d5650d --file-prefix=pf -- -i
190 #. Start the DPDK application that will manage the VF device
192 .. code-block:: console
194 <build_dir>/app/dpdk-testpmd -l 26-29 -n 4 -a 86:02.0 \
195 --vfio-vf-token=14d63f20-8445-11ea-8900-1f9ce7d5650d --file-prefix=vf0 -- -i
197 To make use of full VFIO functionality,
198 both kernel and BIOS must support and be configured
199 to use IO virtualization (such as Intel\ |reg| VT-d).
203 Linux versions earlier than version 3.6 do not support VFIO.
207 Linux versions earlier than version 5.7 do not support the creation of
208 virtual functions within the VFIO framework.
212 In most cases, specifying "iommu=on" as kernel parameter should be enough to
213 configure the Linux kernel to use IOMMU.
215 For proper operation of VFIO when running DPDK applications as a non-privileged user, correct permissions should also be set up.
216 For more information, please refer to :ref:`Running_Without_Root_Privileges`.
223 If there is no IOMMU available on the system, VFIO can still be used,
224 but it has to be loaded with an additional module parameter:
226 .. code-block:: console
228 modprobe vfio enable_unsafe_noiommu_mode=1
230 Alternatively, one can also enable this option in an already loaded kernel module:
232 .. code-block:: console
234 echo 1 > /sys/module/vfio/parameters/enable_unsafe_noiommu_mode
236 After that, VFIO can be used with hardware devices as usual.
240 It may be required to unload all VFIO related-modules before probing
241 the module again with ``enable_unsafe_noiommu_mode=1`` parameter.
245 Since no-IOMMU mode forgoes IOMMU protection, it is inherently unsafe.
246 That said, it does make it possible for the user
247 to keep the degree of device access and programming that VFIO has,
248 in situations where IOMMU is not available.
257 Using UIO drivers is inherently unsafe due to this method lacking IOMMU protection,
258 and can only be done by root user.
260 In situations where using VFIO is not an option, there are alternative drivers one can use.
261 In many cases, the standard ``uio_pci_generic`` module included in the Linux kernel
262 can be used as a substitute for VFIO. This module can be loaded using the command:
264 .. code-block:: console
266 sudo modprobe uio_pci_generic
270 ``uio_pci_generic`` module doesn't support the creation of virtual functions.
272 As an alternative to the ``uio_pci_generic``, there is the ``igb_uio`` module
273 which can be found in the repository `dpdk-kmods <http://git.dpdk.org/dpdk-kmods>`_.
274 It can be loaded as shown below:
276 .. code-block:: console
279 sudo insmod igb_uio.ko
283 For some devices which lack support for legacy interrupts, e.g. virtual function
284 (VF) devices, the ``igb_uio`` module may be needed in place of ``uio_pci_generic``.
288 If UEFI secure boot is enabled,
289 the Linux kernel may disallow the use of UIO on the system.
290 Therefore, devices for use by DPDK should be bound to the ``vfio-pci`` kernel module
291 rather than any UIO-based module.
292 For more details see :ref:`linux_gsg_binding_kernel` below.
296 If the devices used for DPDK are bound to a UIO-based kernel module,
297 please make sure that the IOMMU is disabled or is in passthrough mode.
298 One can add ``intel_iommu=off`` or ``amd_iommu=off`` or ``intel_iommu=on iommu=pt``
299 in GRUB command line on x86_64 systems,
300 or add ``iommu.passthrough=1`` on aarch64 systems.
302 .. _bifurcated_driver:
307 PMDs which use the bifurcated driver co-exists with the device kernel driver.
308 On such model the NIC is controlled by the kernel, while the data
309 path is performed by the PMD directly on top of the device.
311 Such model has the following benefits:
313 - It is secure and robust, as the memory management and isolation
314 is done by the kernel.
315 - It enables the user to use legacy linux tools such as ``ethtool`` or
316 ``ifconfig`` while running DPDK application on the same network ports.
317 - It enables the DPDK application to filter only part of the traffic,
318 while the rest will be directed and handled by the kernel driver.
319 The flow bifurcation is performed by the NIC hardware.
320 As an example, using :ref:`flow_isolated_mode` allows to choose
321 strictly what is received in DPDK.
323 More about the bifurcated driver can be found in
324 `Mellanox Bifurcated DPDK PMD
325 <https://www.dpdk.org/wp-content/uploads/sites/35/2016/10/Day02-Session04-RonyEfraim-Userspace2016.pdf>`__.
330 In certain situations, using ``dpdk-devbind.py`` script
331 to bind a device to VFIO driver may fail.
332 The first place to check is the kernel messages:
334 .. code-block:: console
338 [ 1297.875090] vfio-pci: probe of 0000:31:00.0 failed with error -22
341 In most cases, the ``error -22`` indicates that the VFIO subsystem
342 could not be enabled because there is no IOMMU support.
344 To check whether the kernel has been booted with correct parameters,
345 one can check the kernel command-line:
347 .. code-block:: console
351 Please refer to earlier sections on how to configure kernel parameters
352 correctly for your system.
354 If the kernel is configured correctly, one also has to make sure that
355 the BIOS configuration has virtualization features (such as Intel\ |reg| VT-d).
356 There is no standard way to check if the platform is configured correctly,
357 so please check with your platform documentation to see if it has such features,
358 and how to enable them.
360 In certain distributions, default kernel configuration is such that
361 the no-IOMMU mode is disabled altogether at compile time.
362 This can be checked in the boot configuration of your system:
364 .. code-block:: console
366 cat /boot/config-$(uname -r) | grep NOIOMMU
367 # CONFIG_VFIO_NOIOMMU is not set
369 If ``CONFIG_VFIO_NOIOMMU`` is not enabled in the kernel configuration,
370 VFIO driver will not support the no-IOMMU mode,
371 and other alternatives (such as UIO drivers) will have to be used.