1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2014 Intel Corporation.
7 This chapter describes the packages required to compile the DPDK.
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*.
14 BIOS Setting Prerequisite on x86
15 --------------------------------
17 For the majority of platforms, no special BIOS settings are needed to use basic DPDK functionality.
18 However, for additional HPET timer and power management functionality,
19 and high performance of small packets, BIOS setting changes may be needed.
20 Consult the section on :ref:`Enabling Additional Functionality <Enabling_Additional_Functionality>`
21 for more information on the required changes.
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`.
30 Compilation of the DPDK
31 -----------------------
33 **Required Tools and Libraries:**
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.
42 * coreutils: ``cmp``, ``sed``, ``grep``, ``arch``, etc.
44 * gcc: versions 4.9 or later is recommended for all platforms.
45 On some distributions, some specific compiler flags and linker flags are enabled by
46 default and affect performance (``-fstack-protector``, for example). Please refer to the documentation
47 of your distribution and to ``gcc -dumpspecs``.
49 * libc headers, often packaged as ``gcc-multilib`` (``glibc-devel.i686`` / ``libc6-dev-i386``;
50 ``glibc-devel.x86_64`` / ``libc6-dev`` for 64-bit compilation on Intel architecture;
51 ``glibc-devel.ppc64`` for 64 bit IBM Power architecture;)
53 * Linux kernel headers or sources required to build kernel modules. (kernel - devel.x86_64;
56 * Additional packages required for 32-bit compilation on 64-bit systems are:
58 * glibc.i686, libgcc.i686, libstdc++.i686 and glibc-devel.i686 for Intel i686/x86_64;
60 * glibc.ppc64, libgcc.ppc64, libstdc++.ppc64 and glibc-devel.ppc64 for IBM ppc_64;
64 x86_x32 ABI is currently supported with distribution packages only on Ubuntu
65 higher than 13.10 or recent Debian distribution. The only supported compiler is gcc 4.9+.
67 * libnuma-devel - library for handling NUMA (Non Uniform Memory Access).
69 * Python, version 2.7+ or 3.2+, to use various helper scripts included in the DPDK package.
74 * Intel® C++ Compiler (icc). For installation, additional libraries may be required.
75 See the icc Installation Guide found in the Documentation directory under the compiler installation.
77 * IBM® Advance ToolChain for Powerlinux. This is a set of open source development tools and runtime libraries
78 which allows users to take leading edge advantage of IBM's latest POWER hardware features on Linux. To install
79 it, see the IBM official installation document.
81 * libpcap headers and libraries (libpcap-devel) to compile and use the libpcap-based poll-mode driver.
82 This driver is disabled by default and can be enabled by setting ``CONFIG_RTE_LIBRTE_PMD_PCAP=y`` in the build time config file.
84 * libarchive headers and library are needed for some unit tests using tar to get their resources.
87 Running DPDK Applications
88 -------------------------
90 To run an DPDK application, some customization may be required on the target machine.
97 * Kernel version >= 3.2
99 The kernel version required is based on the oldest long term stable kernel available
100 at kernel.org when the DPDK version is in development.
102 The kernel version in use can be checked using the command::
106 * glibc >= 2.7 (for features related to cpuset)
108 The version can be checked using the ``ldd --version`` command.
110 * Kernel configuration
112 In the Fedora OS and other common distributions, such as Ubuntu, or Red Hat Enterprise Linux,
113 the vendor supplied kernel configurations can be used to run most DPDK applications.
115 For other kernel builds, options which should be enabled for DPDK include:
119 * PROC_PAGE_MONITOR support
121 * HPET and HPET_MMAP configuration options should also be enabled if HPET support is required.
122 See the section on :ref:`High Precision Event Timer (HPET) Functionality <High_Precision_Event_Timer>` for more details.
124 .. _linux_gsg_hugepages:
126 Use of Hugepages in the Linux Environment
127 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
129 Hugepage support is required for the large memory pool allocation used for packet buffers
130 (the HUGETLBFS option must be enabled in the running kernel as indicated the previous section).
131 By using hugepage allocations, performance is increased since fewer pages are needed,
132 and therefore less Translation Lookaside Buffers (TLBs, high speed translation caches),
133 which reduce the time it takes to translate a virtual page address to a physical page address.
134 Without hugepages, high TLB miss rates would occur with the standard 4k page size, slowing performance.
136 Reserving Hugepages for DPDK Use
137 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
139 The allocation of hugepages should be done at boot time or as soon as possible after system boot
140 to prevent memory from being fragmented in physical memory.
141 To reserve hugepages at boot time, a parameter is passed to the Linux kernel on the kernel command line.
143 For 2 MB pages, just pass the hugepages option to the kernel. For example, to reserve 1024 pages of 2 MB, use::
147 For other hugepage sizes, for example 1G pages, the size must be specified explicitly and
148 can also be optionally set as the default hugepage size for the system.
149 For example, to reserve 4G of hugepage memory in the form of four 1G pages, the following options should be passed to the kernel::
151 default_hugepagesz=1G hugepagesz=1G hugepages=4
155 The hugepage sizes that a CPU supports can be determined from the CPU flags on Intel architecture.
156 If pse exists, 2M hugepages are supported; if pdpe1gb exists, 1G hugepages are supported.
157 On IBM Power architecture, the supported hugepage sizes are 16MB and 16GB.
161 For 64-bit applications, it is recommended to use 1 GB hugepages if the platform supports them.
163 In the case of a dual-socket NUMA system,
164 the number of hugepages reserved at boot time is generally divided equally between the two sockets
165 (on the assumption that sufficient memory is present on both sockets).
167 See the Documentation/kernel-parameters.txt file in your Linux source tree for further details of these and other kernel options.
171 For 2 MB pages, there is also the option of allocating hugepages after the system has booted.
172 This is done by echoing the number of hugepages required to a nr_hugepages file in the ``/sys/devices/`` directory.
173 For a single-node system, the command to use is as follows (assuming that 1024 pages are required)::
175 echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages
177 On a NUMA machine, pages should be allocated explicitly on separate nodes::
179 echo 1024 > /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages
180 echo 1024 > /sys/devices/system/node/node1/hugepages/hugepages-2048kB/nr_hugepages
184 For 1G pages, it is not possible to reserve the hugepage memory after the system has booted.
186 On IBM POWER system, the nr_overcommit_hugepages should be set to the same value as nr_hugepages.
187 For example, if the required page number is 128, the following commands are used::
189 echo 128 > /sys/kernel/mm/hugepages/hugepages-16384kB/nr_hugepages
190 echo 128 > /sys/kernel/mm/hugepages/hugepages-16384kB/nr_overcommit_hugepages
192 Using Hugepages with the DPDK
193 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
195 Once the hugepage memory is reserved, to make the memory available for DPDK use, perform the following steps::
198 mount -t hugetlbfs nodev /mnt/huge
200 The mount point can be made permanent across reboots, by adding the following line to the ``/etc/fstab`` file::
202 nodev /mnt/huge hugetlbfs defaults 0 0
204 For 1GB pages, the page size must be specified as a mount option::
206 nodev /mnt/huge_1GB hugetlbfs pagesize=1GB 0 0