QNX Momentics IDE works with the Application Profiler and Valgrind (pronounced “val-grinned”) Cachegrind tools,
which measure the performance of programs running on QNX targets. It also offers the QNX System Information perspective for viewing
performance-related metrics.
Note:
The term performance can have many meanings. In this guide, it refers to application responsiveness and the
speed of operations. Memory consumption, particularly on the heap, is an important part of performance; this topic is covered
in the previous chapter. Here, we discuss performance primarily from a speed perspective.
The analysis tools support these three tasks:
- monitoring performance — you can examine realtime performance statistics for processes and threads
to determine which applications need to be further analyzed and optimized
- profiling applications — you can profile an application to learn how often its individual functions
are called and how much time is spent in each function
- investigating performance bottlenecks — you can measure an application's cache misses and heap memory usage,
and see how it interacts with other processes to determine why it runs slowly
You can export performance analysis data to share results with other
team members, who can then import them into their IDE installations. The import feature also supports postmortem analysis, in which you view analysis results for an application that was
run outside of the IDE.