doc: add missing release notes for QAT

Updated release notes for the changes done in QAT for DPDK 20.02
Original commits for these features are as follows.

Fixes: 026f21c0b951 ("common/qat: support dual threads for enqueue/dequeue")
Fixes: 6cde900bd59d ("common/qat: remove tail write coalescing")

Signed-off-by: Arek Kusztal <arkadiuszx.kusztal@intel.com>
Acked-by: Fiona Trahe <fiona.trahe@intel.com>

diff --git a/doc/guides/rel_notes/release_20_02.rst b/doc/guides/rel_notes/release_20_02.rst
index 786f9b1..c7482c2 100644 (file)
--- a/doc/guides/rel_notes/release_20_02.rst
+++ b/doc/guides/rel_notes/release_20_02.rst
@@ -146,6 +146,13 @@ New Features
   Such algorithm combinations are not supported on GEN1/GEN2 hardware
   and executing the request returns RTE_CRYPTO_OP_STATUS_INVALID_SESSION.
 
   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
 * **Updated the ZUC PMD.**
 
   * Transistioned underlying library from libSSO ZUC to intel-ipsec-mb


@@ -207,6 +214,15 @@ Removed Items
   the Technical Board decided to disable all the kernel modules
   by default from 20.02 version.
 
   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
 -----------
 
 API Changes
 -----------