* ECPM (Elliptic Curve Point Multiplication) is added to
asymmetric crypto library specifications.
+* **Added synchronous Crypto burst API.**
+
+ A new API is introduced in crypto library to handle synchronous cryptographic
+ operations allowing to achieve performance gain for cryptodevs which use
+ CPU based acceleration, such as Intel AES-NI. An implementation for aesni_gcm
+ cryptodev is provided. The IPsec example application and ipsec library itself
+ were changed to allow utilization of this new feature.
+
* **Added handling of mixed algorithms in encrypted digest requests in QAT PMD.**
Added handling of mixed algorithms in encrypted digest hash-cipher
Such algorithm combinations are not supported on GEN1/GEN2 hardware
and executing the request returns RTE_CRYPTO_OP_STATUS_INVALID_SESSION.
+* **Queue-pairs are now thread-safe on Intel QuickAssist Technology (QAT) PMD.**
+
+ Queue-pairs are thread-safe on Intel CPUs but Queues are not (that is, within
+ a single queue-pair all enqueues to the TX queue must be done from one thread
+ and all dequeues from the RX queue must be done from one thread, but enqueues
+ and dequeues may be done in different threads.).
+
* **Updated the ZUC PMD.**
* Transistioned underlying library from libSSO ZUC to intel-ipsec-mb
Add event device support for ``l3fwd`` sample application. It demonstrates
usage of poll and event mode IO mechanism under a single application.
+* **Added cycle-count mode to the compression performance tool.**
+
+ Enhanced the compression performance tool by adding a cycle-count mode
+ which can be used to help measure and tune hardware and software PMDs.
+
Removed Items
-------------
the Technical Board decided to disable all the kernel modules
by default from 20.02 version.
+* **Removed coalescing feature from Intel QuickAssist Technology (QAT) PMD.**
+
+ The internal tail write coalescing feature was removed as not compatible with
+ dual-thread feature. It was replaced with a threshold feature. At busy times
+ if only a small number of packets can be enqueued, each enqueue causes
+ an expensive MMIO write. These MMIO write occurrences can be optimised by using
+ the new threshold parameter on process start. Please see qat documentation for
+ more details.
+
API Changes
-----------