1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2014 Intel Corporation. 3 4.. _Enabling_Additional_Functionality: 5 6Enabling Additional Functionality 7================================= 8 9.. _High_Precision_Event_Timer: 10 11High Precision Event Timer (HPET) Functionality 12----------------------------------------------- 13 14BIOS Support 15~~~~~~~~~~~~ 16 17The High Precision Timer (HPET) must be enabled in the platform BIOS if the HPET is to be used. 18Otherwise, the Time Stamp Counter (TSC) is used by default. 19The BIOS is typically accessed by pressing F2 while the platform is starting up. 20The user can then navigate to the HPET option. On the Crystal Forest platform BIOS, the path is: 21**Advanced -> PCH-IO Configuration -> High Precision Timer ->** (Change from Disabled to Enabled if necessary). 22 23On a system that has already booted, the following command can be issued to check if HPET is enabled:: 24 25 grep hpet /proc/timer_list 26 27If no entries are returned, HPET must be enabled in the BIOS (as per the instructions above) and the system rebooted. 28 29Linux Kernel Support 30~~~~~~~~~~~~~~~~~~~~ 31 32The DPDK makes use of the platform HPET timer by mapping the timer counter into the process address space, and as such, 33requires that the ``HPET_MMAP`` kernel configuration option be enabled. 34 35.. warning:: 36 37 On Fedora, and other common distributions such as Ubuntu, the ``HPET_MMAP`` kernel option is not enabled by default. 38 To recompile the Linux kernel with this option enabled, please consult the distributions documentation for the relevant instructions. 39 40Enabling HPET in the DPDK 41~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 42 43By default, HPET support is disabled in the DPDK build configuration files. 44To use HPET, use the following meson build option which will enable the HPET settings at compile time:: 45 46 meson configure -Duse_hpet=true 47 48For an application to use the ``rte_get_hpet_cycles()`` and ``rte_get_hpet_hz()`` API calls, 49and optionally to make the HPET the default time source for the rte_timer library, 50the new ``rte_eal_hpet_init()`` API call should be called at application initialization. 51This API call will ensure that the HPET is accessible, returning an error to the application if it is not, 52for example, if ``HPET_MMAP`` is not enabled in the kernel. 53The application can then determine what action to take, if any, if the HPET is not available at run-time. 54 55.. note:: 56 57 For applications that require timing APIs, but not the HPET timer specifically, 58 it is recommended that the ``rte_get_timer_cycles()`` and ``rte_get_timer_hz()`` API calls be used instead of the HPET-specific APIs. 59 These generic APIs can work with either TSC or HPET time sources, depending on what is requested by an application call to ``rte_eal_hpet_init()``, 60 if any, and on what is available on the system at runtime. 61 62Running DPDK Applications Without Root Privileges 63-------------------------------------------------------- 64 65.. note:: 66 67 The instructions below will allow running DPDK as non-root with older 68 Linux kernel versions. However, since version 4.0, the kernel does not allow 69 unprivileged processes to read the physical address information from 70 the pagemaps file, making it impossible for those processes to use HW 71 devices which require physical addresses 72 73Although applications using the DPDK use network ports and other hardware resources directly, 74with a number of small permission adjustments it is possible to run these applications as a user other than "root". 75To do so, the ownership, or permissions, on the following Linux file system objects should be adjusted to ensure that 76the Linux user account being used to run the DPDK application has access to them: 77 78* All directories which serve as hugepage mount points, for example, ``/mnt/huge`` 79 80* The userspace-io device files in ``/dev``, for example, ``/dev/uio0``, ``/dev/uio1``, and so on 81 82* The userspace-io sysfs config and resource files, for example for ``uio0``:: 83 84 /sys/class/uio/uio0/device/config 85 /sys/class/uio/uio0/device/resource* 86 87* If the HPET is to be used, ``/dev/hpet`` 88 89.. note:: 90 91 On some Linux installations, ``/dev/hugepages`` is also a hugepage mount point created by default. 92 93Power Management and Power Saving Functionality 94----------------------------------------------- 95 96Enhanced Intel SpeedStep® Technology must be enabled in the platform BIOS if the power management feature of DPDK is to be used. 97Otherwise, the sys file folder ``/sys/devices/system/cpu/cpu0/cpufreq`` will not exist, and the CPU frequency- based power management cannot be used. 98Consult the relevant BIOS documentation to determine how these settings can be accessed. 99 100For example, on some Intel reference platform BIOS variants, the path to Enhanced Intel SpeedStep® Technology is:: 101 102 Advanced 103 -> Processor Configuration 104 -> Enhanced Intel SpeedStep® Tech 105 106In addition, C3 and C6 should be enabled as well for power management. The path of C3 and C6 on the same platform BIOS is:: 107 108 Advanced 109 -> Processor Configuration 110 -> Processor C3 Advanced 111 -> Processor Configuration 112 -> Processor C6 113 114Using Linux Core Isolation to Reduce Context Switches 115----------------------------------------------------- 116 117While the threads used by an DPDK application are pinned to logical cores on the system, 118it is possible for the Linux scheduler to run other tasks on those cores also. 119To help prevent additional workloads from running on those cores, 120it is possible to use the ``isolcpus`` Linux kernel parameter to isolate them from the general Linux scheduler. 121 122For example, if DPDK applications are to run on logical cores 2, 4 and 6, 123the following should be added to the kernel parameter list: 124 125.. code-block:: console 126 127 isolcpus=2,4,6 128 129Loading the DPDK KNI Kernel Module 130---------------------------------- 131 132To run the DPDK Kernel NIC Interface (KNI) sample application, an extra kernel module (the kni module) must be loaded into the running kernel. 133The module is found in the kernel/linux sub-directory of the DPDK build directory. 134It should be loaded using the insmod command:: 135 136 insmod <build_dir>/kernel/linux/kni/rte_kni.ko 137 138.. note:: 139 140 See the "Kernel NIC Interface Sample Application" chapter in the *DPDK Sample Applications User Guide* for more details. 141 142Using Linux IOMMU Pass-Through to Run DPDK with Intel® VT-d 143----------------------------------------------------------- 144 145To enable Intel® VT-d in a Linux kernel, a number of kernel configuration options must be set. These include: 146 147* ``IOMMU_SUPPORT`` 148 149* ``IOMMU_API`` 150 151* ``INTEL_IOMMU`` 152 153In addition, to run the DPDK with Intel® VT-d, the ``iommu=pt`` kernel parameter must be used when using ``igb_uio`` driver. 154This results in pass-through of the DMAR (DMA Remapping) lookup in the host. 155Also, if ``INTEL_IOMMU_DEFAULT_ON`` is not set in the kernel, the ``intel_iommu=on`` kernel parameter must be used too. 156This ensures that the Intel IOMMU is being initialized as expected. 157 158Please note that while using ``iommu=pt`` is compulsory for ``igb_uio driver``, the ``vfio-pci`` driver can actually work with both ``iommu=pt`` and ``iommu=on``. 159