Hello,
I'm working on a custom carrier board mounting an IMX8MP connected with eQOS ethernet interface in RGMII mode to a Marvell 88E6176 switch.
My kernel is a custom NXP based linux 5.15.71 (NO custom modifications on ethernet section) with Yocto Kirkstone rootfs.
Sometimes in the middle of the startup I see IMX8MP ethernet TX clock in RGMII suddenly shutting down with no apparently motivation.
That leads to a hardware reset by watchdog with consequent reboot. This means that some kernel procedure fails and stuck on something.
Is there any reason for microprocessor to act like that?
