Hello Oliver and Carsten and others, we have reached state where we have running CAN latester service automatically again. We use the CTU CAN FD IP core for precise timestamps of messages on a measurement and load generation device and user-space or kernel-space CAN gateway on the device under the tests (DUT, at this moment Zynq based MZ_APO kit with CTU CAN FD as well). But there are minimal requirements on DUT side, device sends all frames received on one CAN interface to another one. All systems supported by SocketCAN can be tested easily and implementation of GW for other systems is simple as well. We have even variant for testing of device with single interface but for flood load it is not ideal setup. The actual results overview with detailed and compare sub-pages https://canbus.pages.fel.cvut.cz/can-latester/ The project is part of CTU CAN/CAN FD projects of FEE at Czech Technical University in Prague https://canbus.pages.fel.cvut.cz/ Our overview last year article about our CAN projects in general and CAN latency testing results and plans in little more details there https://can-newsletter.org/uploads/media/raw/a9abe317ae034be55d99fee4410ad70e.pdf Consider the announced setup as experimental still but we hope that it will mature to durable, long term solution. The actual setup includes Debian x86 virtual dedicated for CAN latency testing control and results processing, a rack with MZ_APO Xilinx Zynq (Linux 6.2.0-rt3 SMP PREEMPT_RT) as measurement/load device, the second MZ_APO Xilinx Zynq as actual/current DUT (mainline and linux-rt-devel for-kbuild-bot/current-stable) kernels and ARM64 sun50i-h5-orangepi-zero-plus (Linux 6.1.15 #5 SMP PREEMPT) PiKRON.com provided testbed controller which controls 24V and 5V power supply and monitors serial ports of the both MZ_APO boards. We plan presentation about GNU/Linux CAN support in general and then about CAN bus/SocketCAN and its realtime capabilities testing at RedHat DevConf CZ (June 17 3:30pm - 4:05pm CEST) https://devconfcz2023.sched.com/event/1MYjG/can-bus-in-control-automotive-and-satellitelites We will be happy for feedback and proposals for future development directions. But be prepared, that we are slow, I spent the most of my time with computer architectures teaching these days and I have some new motion control project at company related to minor contribution to the ground experiments in the frame of LISA (concept of low-frequency gravitational wave detection in space) project at company. But CAN in Linux, NuttX and RTEMS kernels and in FPGAs is area of my long term interrest. I hope that we reach mainline with CTU CAN FD timestamping patches. As for actual setup, we see problems with preemp RT kernel on Zynq platform with 6.3+ kernels (6.2 and previous versions has been stable for years), but the issue seems to be unrelated to SocketCAN so it is discussion for another list. I want to thanks to colleagues, our graduates and studnets who have contributed to the project and made it possible. The theses directly related to the project https://dspace.cvut.cz/bitstream/handle/10467/80366/F3-DP-2019-Jerabek-Martin-Jerabek-thesis-2019-canfd.pdf https://dspace.cvut.cz/bitstream/handle/10467/101450/F3-DP-2022-Vasilevski-Matej-vasilmat.pdf Pavel Hronek is now finishing his own bachelor thesis related mainly to the CAN latency testing automation and web presentation, it will be available after submition on the list our CAN theses at https://canbus.pages.fel.cvut.cz/ as well as the wider reach of theses of OTREES informal group https://gitlab.fel.cvut.cz/otrees/org/-/wikis/theses-defend Best wishes, Pavel Pisa phone: +420 603531357 e-mail: pisa@xxxxxxxxxxxxxxxx Department of Control Engineering FEE CVUT Karlovo namesti 13, 121 35, Prague 2 university: http://control.fel.cvut.cz/ personal: http://cmp.felk.cvut.cz/~pisa projects: https://www.openhub.net/accounts/ppisa CAN related:http://canbus.pages.fel.cvut.cz/ RISC-V education: https://comparch.edu.cvut.cz/ Open Technologies Research Education and Exchange Services https://gitlab.fel.cvut.cz/otrees/org/-/wikis/home
