memory that was created outside of DPDK's own page allocator, and using that
memory natively with any other DPDK library or data structure.
+* **Added check for ensuring allocated memory addressable by devices.**
+
+ Some devices can have addressing limitations so a new function,
+ ``rte_mem_check_dma_mask``, has been added for checking allocated memory is
+ not out of the device range. Because now memory can be dynamically allocated
+ after initialization, a dma mask is kept and any new allocated memory will be
+ checked out against that dma mask and rejected if out of range. If more than
+ one device has addressing limitations, the dma mask is the more restricted one.
+
+* **Updated the C11 memory model version of ring library.**
+
+ The latency is decreased for architectures using the C11 memory model
+ version of the ring library.
+
+ On Cavium ThunderX2 platform, the changes decreased latency by 27~29%
+ and 3~15% for MPMC and SPSC cases respectively (with 2 lcores). The
+ real improvements may vary with the number of contending lcores and
+ the size of ring.
+
* **Added hot-unplug handle mechanism.**
``rte_dev_hotplug_handle_enable`` and ``rte_dev_hotplug_handle_disable`` are
See the :doc:`../nics/atlantic` nic driver guide for more details on this
driver.
+* **Updated mlx5 driver.**
+
+ Updated the mlx5 driver including the following changes:
+
+ * Improved security of PMD to prevent the NIC from getting stuck when
+ application misbehaves.
+ * reworked flow engine to supported e-switch flow rules (transfer attribute).
+ * Supported header re-write(L2-L4), VXLAN encap/decap, count ,match
+ on TCP flags and multiple flow groups with e-switch flow rules.
+ * Supported match on medatada, VXLAN and MPLS encap/decap with flow rules.
+ * Supported RTE_ETH_DEV_CLOSE_REMOVE flag to provide better support for
+ representors.
+ * Supported meson build.
+ * Fixed build issue with PPC.
+ * Supported BlueField VF.
+ * Supported externally allocated static memory for DMA.
+
* **Updated Solarflare network PMD.**
Updated the sfc_efx driver including the following changes:
* Added support for Tx descriptor status API in EF10 (ef10 and ef10_simple)
datapaths implementation.
+* **Updated the enic driver.**
+
+ * Added AVX2-based vectorized Rx handler.
+ * Added VLAN and checksum offloads to the simple Tx handler.
+ * Added the count flow action.
+ * Enabled the virtual address IOVA mode.
+
* **Updated failsafe driver.**
Updated the failsafe driver including the following changes:
A new devarg ``phy_mac`` was introduced to allow users to use physical
MAC address of the selected PCAP interface.
+* **Added TAP Rx/Tx queues sharing with a secondary process.**
+
+ A secondary process can attach a TAP device created in the primary process,
+ probe the queues, and process Rx/Tx in a secondary process.
+
+* **Added classification and metering support to SoftNIC PMD.**
+
+ Added support for flow classification (rte_flow API), and metering and
+ policing (rte_mtr API) to the SoftNIC PMD.
+
+* **Added Crypto support to Softnic PMD.**
+
+ The Softnic is now capable of processing symmetric crypto workloads such
+ as cipher, cipher-authentication chaining, and aead encryption and
+ decryption. This is achieved by calling DPDK Cryptodev APIs.
+
+* **Added cryptodev port to port library.**
+
+ Cryptodev port is a shim layer in the port library that interacts with DPDK
+ Cryptodev PMDs including burst enqueuing and dequeuing crypto operations.
+
+* **Added symmetric cryptographic actions to the pipeline library.**
+
+ In the pipeline library an added symmetric crypto action parsing and action
+ handler are implemented. The action allows automatically preparing the crypto
+ operation with the rules specified such as algorithm, key, and IV, etc for
+ the cryptodev port to process.
+
* **Added support for GEN3 devices to Intel QAT driver .**
Added support for the third generation of Intel QuickAssist devices.
Added the new caam job ring driver for NXP platforms. See the
"NXP CAAM JOB RING (caam_jr)" document for more details on this new driver.
+* **Added support for Dynamic Huffman Encoding to Intel QAT comp PMD.**
+
+ The Intel QuickAssist (QAT) compression PMD has been updated with support
+ for Dynamic Huffman Encoding for the Deflate algorithm.
+
* **Added Event Ethernet Tx Adapter.**
Added event ethernet Tx adapter library that provides configuration and
the SW eventdev PMD, sacrifices load balancing performance to
gain better event scheduling throughput and scalability.
+* **Added extendable bucket feature to hash library (rte_hash).**
+
+ This new “extendable bucket” feature provides 100% insertion guarantee to
+ the capacity specified by the user by extending hash table with extra
+ buckets when needed to accommodate the unlikely event of intensive hash
+ collisions. In addition, the internal hashing algorithm was changed to use
+ partial-key hashing to improve memory efficiency and lookup performance.
+
+* **Added lock free reader/writer concurrency to hash library (rte_hash).**
+
+ Lock free reader/writer concurrency prevents the readers from getting
+ blocked due to a pre-empted writer thread. This allows the hash library
+ to be used in scenarios where the writer thread runs on control plane.
+
+* **Added Traffic Pattern Aware Power Control Library**
+
+ Added an experimental library. This extend Power Library and provide
+ empty_poll APIs. This feature measure how many times empty_poll are
+ executed per core, use the number of empty polls as a hint for system
+ power management.
+
+ See the :doc:`../prog_guide/power_man` section of the DPDK Programmers
+ Guide document for more information.
+
+* **Added JSON power policy interface for containers.**
+
+ Extended the Power Library and vm_power_manager sample app to allow power
+ policies to be submitted via a FIFO using JSON formatted strings. Previously
+ limited to Virtual Machines, this feature extends power policy functionality
+ to containers and host applications that need to have their cores frequency
+ controlled based on the rules contained in the policy.
+
+* **Added Telemetry API.**
+
+ Added the telemetry API which allows applications to transparently expose
+ their telemetry via a UNIX socket in JSON. The JSON can be consumed by any
+ Service Assurance agent, such as CollectD.
+
+* **Updated KNI kernel module, rte_kni library, and KNI sample application.**
+
+ Updated the KNI kernel module with a new kernel module parameter,
+ ``carrier=[on|off]`` to allow the user to control the default carrier
+ state of KNI kernel network interfaces. The default carrier state
+ is now set to ``off``, so the interfaces cannot be used until the
+ carrier state is set to ``on`` via ``rte_kni_update_link`` or
+ by writing ``1`` to ``/sys/devices/virtual/net/<iface>/carrier``.
+ In previous versions the default carrier state was left undefined.
+ See :doc:`../prog_guide/kernel_nic_interface` for more information.
+
+ Added the new API function ``rte_kni_update_link`` to allow the user
+ to set the carrier state of the KNI kernel network interface.
+
+ Added a new command line flag ``-m`` to the KNI sample application to
+ monitor and automatically reflect the physical NIC carrier state to the
+ KNI kernel network interface with the new ``rte_kni_update_link`` API.
+ See :doc:`../sample_app_ug/kernel_nic_interface` for more information.
+
* **Added ability to switch queue deferred start flag on testpmd app.**
Added a console command to testpmd app, giving ability to switch
this application doesn't need to launch dedicated worker threads for vhost
enqueue/dequeue operations.
+* **Added cryptodev FIPS validation example application.**
+
+ Added an example application to parse and perform symmetric cryptography
+ computation to the NIST Cryptographic Algorithm Validation Program (CAVP)
+ test vectors.
+
+* **Allow unit test binary to take parameters from the environment**
+
+ The unit test "test", or "dpdk-test", binary is often called from scripts,
+ which can make passing additional parameters, such as a coremask, to it more
+ awkward. Support has been added to the application to allow it to take
+ additional command-line parameter values from the "DPDK_TEST_PARAMS"
+ environment variable to make this application easier to use.
+
API Changes
-----------
* eal: The parameters of the function ``rte_devargs_remove()`` have changed
from bus and device names to ``struct rte_devargs``.
+* eal: The deprecated functions attach/detach were removed in 18.11.
+ ``rte_eal_dev_attach`` can be replaced by
+ ``rte_dev_probe`` or ``rte_eal_hotplug_add``.
+ ``rte_eal_dev_detach`` can be replaced by
+ ``rte_dev_remove`` or ``rte_eal_hotplug_remove``.
+
* eal: The scope of ``rte_eal_hotplug_add()``/``rte_dev_probe()``
and ``rte_eal_hotplug_remove()``/``rte_dev_remove()`` is extended.
In multi-process model, they will guarantee that the device is
functions were deprecated since 17.05 and are replaced by
``rte_mbuf_raw_free()`` and ``rte_pktmbuf_prefree_seg()``.
+* ethdev: The deprecated functions attach/detach were removed in 18.11.
+ ``rte_eth_dev_attach`` can be replaced by ``RTE_ETH_FOREACH_MATCHING_DEV``
+ and ``rte_dev_probe`` or ``rte_eal_hotplug_add``.
+ ``rte_eth_dev_detach`` can be replaced by
+ ``rte_dev_remove`` or ``rte_eal_hotplug_remove``.
+
+* ethdev: A call to ``rte_eth_dev_release_port()`` has been added in
+ ``rte_eth_dev_close()``. As a consequence, a closed port is freed
+ and seen as invalid because of its state ``RTE_ETH_DEV_UNUSED``.
+ This new behaviour is enabled per driver for a migration period.
+
* A new device flag, RTE_ETH_DEV_NOLIVE_MAC_ADDR, changes the order of
actions inside rte_eth_dev_start regarding MAC set. Some NICs do not
support MAC changes once the port has started and with this new device
* eventdev: Type of 2nd parameter to ``rte_event_eth_rx_adapter_caps_get()``
has been changed from uint8_t to uint16_t.
+* kni: By default, interface carrier status is ``off`` which means there won't
+ be any traffic. It can be set to ``on`` via ``rte_kni_update_link()`` API
+ or via ``sysfs`` interface:
+ ``echo 1 > /sys/class/net/vEth0/carrier``.
+ Note interface should be ``up`` to be able to read/write sysfs interface.
+ When KNI sample application is used, ``-m`` parameter can be used to
+ automatically update the carrier status for the interface.
+
+* kni: When ethtool support enabled (``CONFIG_RTE_KNI_KMOD_ETHTOOL=y``)
+ ethtool commands ``ETHTOOL_GSET & ETHTOOL_SSET`` are no more supported for the
+ kernels that has ``ETHTOOL_GLINKSETTINGS & ETHTOOL_SLINKSETTINGS`` support.
+ This means ``ethtool "-a|--show-pause", "-s|--change"`` won't work, and
+ ``ethtool <iface>`` output will have less information.
+
ABI Changes
-----------
- structure ``rte_eal_memconfig`` has been extended to contain next
socket ID for externally allocated segments
+* eal: Added ``dma_maskbits`` to ``rte_mem_config`` for keeping more restricted
+ dma mask based on devices addressing limitations.
+
* eal: The structure ``rte_device`` got a new field to reference a ``rte_bus``.
It is changing the size of the ``struct rte_device`` and the inherited
device structures of all buses.
+ librte_bus_vmbus.so.2
librte_cfgfile.so.2
librte_cmdline.so.2
- librte_common_octeontx.so.1
librte_compressdev.so.1
librte_cryptodev.so.5
librte_distributor.so.1
+ librte_eal.so.9
+ librte_efd.so.1
+ librte_ethdev.so.11
+ librte_eventdev.so.6
librte_flow_classify.so.1
librte_latencystats.so.1
librte_lpm.so.2
librte_mbuf.so.4
+ librte_member.so.1
librte_mempool.so.5
librte_meter.so.2
librte_metrics.so.1
librte_pmd_bond.so.2
librte_pmd_i40e.so.2
librte_pmd_ixgbe.so.2
- librte_pmd_dpaa2_cmdif.so.1
librte_pmd_dpaa2_qdma.so.1
librte_pmd_ring.so.2
librte_pmd_softnic.so.1
librte_pmd_vhost.so.2
- + librte_pmd_netvsc.so.1
librte_port.so.3
librte_power.so.1
librte_rawdev.so.1
librte_security.so.1
librte_table.so.3
librte_timer.so.1
- librte_vhost.so.3
+ librte_vhost.so.4
Known Issues
driver; the Linux netvsc device must be brought up before the netvsc device is
unbound and passed to the DPDK.
+* IBM Power8 is not supported by this release of DPDK. IBM Power9 is supported.
+
+* ``AVX-512`` support has been disabled for ``GCC`` builds [1] because of a crash [2].
+ This can affect ``native`` machine type build targets on the platforms that support
+ ``AVX512F`` like ``Intel Skylake`` processors, and can cause a possible performance drop.
+ The immediate workaround is to use ``clang`` compiler on these platforms.
+ Issue has been identified as a GCC defect and reported to GCC community [3]. Further
+ actions will be taken based on the GCC defect result.
+
+ [1]: Commit 8d07c82b239f ("mk: disable gcc AVX512F support")
+ [2]: https://bugs.dpdk.org/show_bug.cgi?id=97
+ [3]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88096
+
Tested Platforms
----------------
git.droids-corp.org / dpdk.git / blobdiff