With the JetPack 4.2.1 we see random CPU sErrors with our custom carrier board. The corresponding log is attached. Sometimes the system boots normally without error.
In JetPack 3.3, we never saw this errors.
What could be the cause?
log.txt (24.2 KB)
Do you have any peripheral on your device when hit this error?
Connected is a display over display port and an I210 Gb Ethernet NIC is connected to a x1 PCIe Interface.
Would you mind removing I210 NIC and see if issue is still there? We need to narrow down the cause first.
We narrowed the problem down. It seems that the pinmux config creates the problem. Can you confirm that the new pinmux sheet from 15.04.2019 Revision 1.00 has to be used with JetPack 4.2.2 and that the older spreadsheet from 02.11.2018 Revision 1.08 cannot be used?
Yes, please use the latest pinmux first. We also received some bug report about old pinmux. Though they are not cpu error.
We are using the latest pinmux sheet now, but still face some problems.
When we do not change anything in the sheet and create the dtsi files and from them the cfg files, the system is booting very slowly (~3min) and also reacting very slow.
If we use the tegra186-mb1-bct-pinmux-quill-p3310-1000-c03.cfg, the system boots normally and the system works without any problems.
Is the default configuration of the Spreadsheet not the same like the tegra186-mb1-bct-pinmux-quill-p3310-1000-c03.cfg? Checking the generated file with the reference, it seems not to be. What is changed?
At least we do not see the kernel panics anymore.
Thank you for your help.
Any news on the topic? With our own pinmux configuration, the system is also very slow.
What could be the reason?
We need a solution for this problem, as we cannot work with our custom board and the standard pinmux file.
It seemed to be an error in the Pinmux Template Sheet. With Version 1.02, it is working now.