Power Framework Benchmarks

Power Framework currently has microbenchmarks exercising the TakeWakeLease call from the System Activity Governor (SAG), and the Lease operation fidls from the Topology Test Daemon. These benchmarks are based on the Criterion benchmark infrastructure. Benchmark functions and fidl proxy obtaining functions are separately declared in each some_work.rs and can be run either as standard integration tests or to profile the performance of the corresponding fidl. The integration tests run in CQ and can verify the correctness of the implementations of the same function for fidl proxy creation and benchmarks.

Running the Benchmarks

1. Add the benchmark test target to your fx set line, and configure your build for release (optimized). Note that the benchmarks rely on SL4F which is currently only available on terminal and workstation builds:

   fx set terminal.x64 --with //src/tests/end_to_end/perf:test --release
     --with-test //src/power:tests #integration test will be included
   

2. Build Fuchsia

   fx build
   

3. Start the Fuchsia emulator

   ffx emu start --headless --net tap
   

4. In a separate terminal, serve Fuchsia packages

   fx serve -v
   

5. Run the tests

   fx test --e2e power_framework_microbenchmarks -o
   

6. Run integration tests (optional)

   fx test -o power-framework-bench-integration-tests --test-filter=*test_topologytestdaemon_toggle -- --repeat 2000
   

After completing, the tests will print the name of the catapult_json output file containing the benchmark results.