1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2014 Intel Corporation. 3 4System Requirements 5=================== 6 7This chapter describes the packages required to compile the DPDK. 8 9.. note:: 10 11 If the DPDK is being used on an Intel® Communications Chipset 89xx Series platform, 12 please consult the *Intel® Communications Chipset 89xx Series Software for Linux Getting Started Guide*. 13 14BIOS Setting Prerequisite on x86 15-------------------------------- 16 17For the majority of platforms, no special BIOS settings are needed to use basic DPDK functionality. 18However, for additional HPET timer and power management functionality, 19and high performance of small packets, BIOS setting changes may be needed. 20Consult the section on :ref:`Enabling Additional Functionality <Enabling_Additional_Functionality>` 21for more information on the required changes. 22 23.. note:: 24 25 If UEFI secure boot is enabled, the Linux kernel may disallow the use of 26 UIO on the system. Therefore, devices for use by DPDK should be bound to the 27 ``vfio-pci`` kernel module rather than ``igb_uio`` or ``uio_pci_generic``. 28 For more details see :ref:`linux_gsg_binding_kernel`. 29 30Compilation of the DPDK 31----------------------- 32 33**Required Tools and Libraries:** 34 35.. note:: 36 37 The setup commands and installed packages needed on various systems may be different. 38 For details on Linux distributions and the versions tested, please consult the DPDK Release Notes. 39 40* General development tools including a supported C compiler such as gcc (version 4.9+) or clang (version 3.4+). 41 42 * For RHEL/Fedora systems these can be installed using ``dnf groupinstall "Development Tools"`` 43 * For Ubuntu/Debian systems these can be installed using ``apt install build-essential`` 44 * For Alpine Linux, ``apk add gcc libc-dev bsd-compat-headers libexecinfo-dev`` 45 46* Python 3.5 or later. 47 48* Meson (version 0.49.2+) and ninja 49 50 * ``meson`` & ``ninja-build`` packages in most Linux distributions 51 * If the packaged version is below the minimum version, the latest versions 52 can be installed from Python's "pip" repository: ``pip3 install meson ninja`` 53 54* ``pyelftools`` (version 0.22+) 55 56 * For Fedora systems it can be installed using ``dnf install python-pyelftools`` 57 * For RHEL/CentOS systems it can be installed using ``pip3 install pyelftools`` 58 * For Ubuntu/Debian it can be installed using ``apt install python3-pyelftools`` 59 * For Alpine Linux, ``apk add py3-elftools`` 60 61* Library for handling NUMA (Non Uniform Memory Access). 62 63 * ``numactl-devel`` in RHEL/Fedora; 64 * ``libnuma-dev`` in Debian/Ubuntu; 65 * ``numactl-dev`` in Alpine Linux 66 67.. note:: 68 69 Please ensure that the latest patches are applied to third party libraries 70 and software to avoid any known vulnerabilities. 71 72 73**Optional Tools:** 74 75* Intel® C++ Compiler (icc). For installation, additional libraries may be required. 76 See the icc Installation Guide found in the Documentation directory under the compiler installation. 77 78* IBM® Advance ToolChain for Powerlinux. This is a set of open source development tools and runtime libraries 79 which allows users to take leading edge advantage of IBM's latest POWER hardware features on Linux. To install 80 it, see the IBM official installation document. 81 82**Additional Libraries** 83 84A number of DPDK components, such as libraries and poll-mode drivers (PMDs) have additional dependencies. 85For DPDK builds, the presence or absence of these dependencies will be automatically detected 86enabling or disabling the relevant components appropriately. 87 88In each case, the relevant library development package (``-devel`` or ``-dev``) is needed to build the DPDK components. 89 90For libraries the additional dependencies include: 91 92* libarchive: for some unit tests using tar to get their resources. 93 94* libelf: to compile and use the bpf library. 95 96For poll-mode drivers, the additional dependencies for each driver can be 97found in that driver's documentation in the relevant DPDK guide document, 98e.g. :doc:`../nics/index` 99 100 101Building DPDK Applications 102-------------------------- 103 104The tool pkg-config or pkgconf, integrated in most build systems, 105must be used to parse options and dependencies from libdpdk.pc. 106 107.. note:: 108 109 pkg-config 0.27, supplied with RHEL-7, 110 does not process the Libs.private section correctly, 111 resulting in statically linked applications not being linked properly. 112 113 114Running DPDK Applications 115------------------------- 116 117To run a DPDK application, some customization may be required on the target machine. 118 119System Software 120~~~~~~~~~~~~~~~ 121 122**Required:** 123 124* Kernel version >= 4.4 125 126 The kernel version required is based on the oldest long term stable kernel available 127 at kernel.org when the DPDK version is in development. 128 Compatibility for recent distribution kernels will be kept, notably RHEL/CentOS 7. 129 130 The kernel version in use can be checked using the command:: 131 132 uname -r 133 134* glibc >= 2.7 (for features related to cpuset) 135 136 The version can be checked using the ``ldd --version`` command. 137 138* Kernel configuration 139 140 In the Fedora OS and other common distributions, such as Ubuntu, or Red Hat Enterprise Linux, 141 the vendor supplied kernel configurations can be used to run most DPDK applications. 142 143 For other kernel builds, options which should be enabled for DPDK include: 144 145 * HUGETLBFS 146 147 * PROC_PAGE_MONITOR support 148 149 * HPET and HPET_MMAP configuration options should also be enabled if HPET support is required. 150 See the section on :ref:`High Precision Event Timer (HPET) Functionality <High_Precision_Event_Timer>` for more details. 151 152.. _linux_gsg_hugepages: 153 154Use of Hugepages in the Linux Environment 155~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 156 157Hugepage support is required for the large memory pool allocation used for packet buffers 158(the HUGETLBFS option must be enabled in the running kernel as indicated the previous section). 159By using hugepage allocations, performance is increased since fewer pages are needed, 160and therefore less Translation Lookaside Buffers (TLBs, high speed translation caches), 161which reduce the time it takes to translate a virtual page address to a physical page address. 162Without hugepages, high TLB miss rates would occur with the standard 4k page size, slowing performance. 163 164Reserving Hugepages for DPDK Use 165^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 166 167The reservation of hugepages can be performed at run time. 168This is done by echoing the number of hugepages required 169to a ``nr_hugepages`` file in the ``/sys/kernel/`` directory 170corresponding to a specific page size (in Kilobytes). 171For a single-node system, the command to use is as follows 172(assuming that 1024 of 2MB pages are required):: 173 174 echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages 175 176On a NUMA machine, the above command will usually divide the number of hugepages 177equally across all NUMA nodes (assuming there is enough memory on all NUMA nodes). 178However, pages can also be reserved explicitly on individual NUMA nodes 179using a ``nr_hugepages`` file in the ``/sys/devices/`` directory:: 180 181 echo 1024 > /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages 182 echo 1024 > /sys/devices/system/node/node1/hugepages/hugepages-2048kB/nr_hugepages 183 184The tool ``dpdk-hugepages.py`` can be used to manage hugepages. 185 186.. note:: 187 188 Some kernel versions may not allow reserving 1 GB hugepages at run time, 189 so reserving them at boot time may be the only option. 190 Please see below for instructions. 191 192**Alternative:** 193 194In the general case, reserving hugepages at run time is perfectly fine, 195but in use cases where having lots of physically contiguous memory is required, 196it is preferable to reserve hugepages at boot time, 197as that will help in preventing physical memory from becoming heavily fragmented. 198 199To reserve hugepages at boot time, a parameter is passed to the Linux kernel on the kernel command line. 200 201For 2 MB pages, just pass the hugepages option to the kernel. For example, to reserve 1024 pages of 2 MB, use:: 202 203 hugepages=1024 204 205For other hugepage sizes, for example 1G pages, the size must be specified explicitly and 206can also be optionally set as the default hugepage size for the system. 207For example, to reserve 4G of hugepage memory in the form of four 1G pages, the following options should be passed to the kernel:: 208 209 default_hugepagesz=1G hugepagesz=1G hugepages=4 210 211.. note:: 212 213 The hugepage sizes that a CPU supports can be determined from the CPU flags on Intel architecture. 214 If pse exists, 2M hugepages are supported; if pdpe1gb exists, 1G hugepages are supported. 215 On IBM Power architecture, the supported hugepage sizes are 16MB and 16GB. 216 217.. note:: 218 219 For 64-bit applications, it is recommended to use 1 GB hugepages if the platform supports them. 220 221In the case of a dual-socket NUMA system, 222the number of hugepages reserved at boot time is generally divided equally between the two sockets 223(on the assumption that sufficient memory is present on both sockets). 224 225See the Documentation/admin-guide/kernel-parameters.txt file in your Linux source tree for further details of these and other kernel options. 226 227Using Hugepages with the DPDK 228^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 229 230If secondary process support is not required, DPDK is able to use hugepages 231without any configuration by using "in-memory" mode. 232Please see :doc:`linux_eal_parameters` for more details. 233 234If secondary process support is required, 235mount points for hugepages need to be created. 236On modern Linux distributions, a default mount point for hugepages 237is provided by the system and is located at ``/dev/hugepages``. 238This mount point will use the default hugepage size 239set by the kernel parameters as described above. 240 241However, in order to use hugepage sizes other than the default, it is necessary 242to manually create mount points for those hugepage sizes (e.g. 1GB pages). 243 244To make the hugepages of size 1GB available for DPDK use, 245following steps must be performed:: 246 247 mkdir /mnt/huge 248 mount -t hugetlbfs pagesize=1GB /mnt/huge 249 250The mount point can be made permanent across reboots, by adding the following line to the ``/etc/fstab`` file:: 251 252 nodev /mnt/huge hugetlbfs pagesize=1GB 0 0 253