We collect benchmarks first by tracing a given process, and then by post-processing the trace to output benchmark information.
There is only one benchmark here currently but more will be added.
This script collects a trace of hello_scenic. It takes one parameter, which is where the outputted benchmarks JSON file should be written.
A Go program that takes the following parameters in this order:
fx shell "/pkgfs/packages/scenic_benchmarks/0/bin/gfx_benchmarks_standalone.sh /tmp \ --catapult-converter-args \ --bots fake_bot_name \ --masters fuchsia.try \ --execution-timestamp-ms 1534438419097 \ --log-url https://fake_url"