When an embedded device crashes in the field, traditional debugging with an attached debugger is often impossible. Coredumps solve this by capturing CPU registers, stack contents, thread state, and variable values at the moment of the crash. This walkthrough covers configuring Zephyr RTOS coredump support on the FRDM-MCXN947 board: enabling the coredump subsystem via Kconfig, adding a Devicetree overlay to reserve a dedicated flash partition, integrating the Spotflow SDK, and uploading the firmware's .elf symbol file. Once configured, Spotflow automatically captures and uploads crash data after reboot, enabling remote post-mortem analysis including stack traces, register inspection, variable state, and AI-assisted crash summaries.
Table of contents
What exactly is a Coredump?Intentionally Crashing the DeviceThe First AttemptBuilding the Application AgainAnalyzing the CrashConclusion59 Impressions