Trace-Based Debugging

Overview

System-wide event tracing using ITM/SWO, ETM, J-Link RTT, SEGGER SystemView, Percepio Tracealyzer, or LTTng to capture timing, task switches, ISRs, and instrumentation events with minimal intrusion. Replaces ad-hoc printf for performance, scheduling, and concurrency debugging.

Benefits

• Captures real-time behavior without disrupting timing budgets
• Visualizes scheduler, ISR, and resource interactions across the whole system
• Far better signal-to-noise than printf-style instrumentation
• Same instrumentation can stream from production via RTT or USB

Limitations & Risks

• Requires hardware support (SWO/ETM) or RTT for high-rate streams
• Tooling is vendor-specific (SEGGER, Lauterbach, Percepio) and can be expensive
• Adoption requires engineers to change long-held debugging habits

Recommended Actions

Standardize on one tracing stack (SystemView + RTT or Tracealyzer + RTOS hooks), instrument key task and ISR boundaries as part of every module's bring-up, and make trace files a normal artifact attached to bug reports.

Additional Notes

Directly displaces the Printf-based Debugging entry (Deprecate). Pairs with observability platforms like Memfault for production-side equivalents.

References & Links

• https://www.segger.com/products/development-tools/systemview/
• https://percepio.com/tracealyzer/
• https://lttng.org/