Skip to content

Commit d1873cc

Browse files
h-nagallapraneethbajjuri
authored andcommitted
feat(linux): AM64X: update 12.0 performance numbers for IPC
update RPMessage/IPC latency performance numbers for 12.0 release. Signed-off-by: Hari Nagalla <hnagalla@ti.com>
1 parent 48ecb86 commit d1873cc

5 files changed

Lines changed: 38 additions & 0 deletions

File tree

source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst

Lines changed: 38 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -646,3 +646,41 @@ Listed for each algorithm are the code snippets used to run each benchmark test.
646646
time -v openssl speed -elapsed -evp aes-128-cbc
647647

648648
|
649+
650+
RP Message Inter-Processor Communication (IPC) Latency
651+
-------------------------------------------------------
652+
653+
RP Message latency Performance
654+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
655+
656+
RP Message latency is the delay measured from sending a round trip echo message from
657+
a Linux application to a remote processor and back. The following measurements use
658+
a RP message length of 1 byte and 490 bytes for comparison.
659+
660+
The Linux user space application `rpmsg_char_benchmark <https://git.ti.com/cgit/rpmsg/ti-rpmsg-char/>`__ captures these latency values.
661+
662+
Test commands used for running IPC latency tests:
663+
664+
.. code:: console
665+
666+
rpmsg_char_benchmark -r 2 -n 100000 -m 1 & chrt -f -p 80 $!
667+
668+
Latencies reported:
669+
670+
.. csv-table::
671+
:header: "Remote Processor","Message Size (in bytes)","Average round trip (usecs)","Max round trip (usecs)"
672+
673+
"R5F0_0","1","35","194"
674+
"R5F0_0","490","154","255"
675+
"M4F","1","43","162"
676+
"M4F","490","282","439"
677+
678+
.. image:: img/R5_0_0_m1_ipc_latency.png
679+
680+
.. image:: img/R5_0_0_m490_ipc_latency.png
681+
682+
.. image:: img/m4_0_m1_ipc_latency.png
683+
684+
.. image:: img/m4_0_m490_ipc_latency.png
685+
686+
|
-355 Bytes
Loading
611 Bytes
Loading
78 Bytes
Loading
-189 Bytes
Loading

0 commit comments

Comments
 (0)