Debugging 64-bit Multi-Core Applications
By Mark Hermeling
As more organizations implement devices based on 64-bit multi-core architectures, Wind River Workbench is designed to make even the most difficult software projects less challenging. Debugging using Workbench helps you find and fix problems in code early on, accelerate application development, reduce device code complexity, improve code quality, and shorten time spent in test cycles.
Supporting Wind River’s 64-bit VxWorks and Linux offerings, Workbench’s Run-Time Analysis tools become especially critical when using a 64-bit multi-core processor such as the NXP T2080 found on boards like Curtiss Wright’s VPX3-133 or VPX6-195. The NXP T2080 has four dual-threaded cores that allow for multiple existing applications to be combined on a single board saving you space, power and cost.
Workbench’s Run-time Analysis tools provide you with visibility into the entire platform: application code, third-party libraries, and the operating system, and most importantly in our example above for multiple cores, tasks, processes, threads and physical processors. With Workbench’s CPU Profiler, you can view all of the threads and tasks aggregated by core CPU (8 virtual cores are shown in our demo video below) or you can deep dive into a specific core’s threads and tasks for a more detailed view on its activity (kernel memory, interrupts, etc.).
A short demonstration using Wind River ’s Workbench Run-Time Analysis tools to debug example 64-bit application code with VxWorks running in 64-bit mode on a Curtiss Wright VPX3-133 board (containing NXP‘s QorIQ T2080 4 dual-threaded 64-bit Power Architecture cores) is available in the following video here.
For more information on Workbench Run-Time Analysis tools, please contact us.
For more information about migrating applications to 64-bit and multi-core, you can refer to our previous blog Scaling your Power Architecture Applications.