You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
67 lines
2.6 KiB
67 lines
2.6 KiB
# This experiment template shows how to run Telemetry tests (using autotest)
|
|
# with explicitly specified DUT configurations.
|
|
#
|
|
# You should replace all the placeholders, marked by angle-brackets,
|
|
# with the appropriate actual values.
|
|
|
|
name: dut_config_telemetry_crosperf_example
|
|
board: <your-board-goes-here>
|
|
|
|
# Note: You can specify multiple remotes, to run your tests in parallel on
|
|
# multiple machines. e.g. "remote: test-machine-1.com test-machine2.come
|
|
# test-machine3.com"
|
|
remote: <your-remote-goes-here>
|
|
|
|
# DUT configuration parameters. All are optional.
|
|
#
|
|
# Run turbostat process in background. Default: True.
|
|
turbostat: <True|False>
|
|
# Run top process in background with specified interval of sampling in
|
|
# seconds, type float. 0 - don't run top.
|
|
# Default: 0
|
|
# Recommended values 1-5 (Lower number provides more accurate data).
|
|
# NOTE: Running top with interval 1-5 sec has insignificant
|
|
# performance impact (performance degradation does not exceed 0.3%).
|
|
top_interval: <interval_in_seconds_float>
|
|
# One of Intel Pstate modes defined in kernel command line:
|
|
# active, passive, no_hwp.
|
|
intel_pstate: <active|passive|no_hwp>
|
|
# Wait until CPU cools down to a specified temperature
|
|
# in Celsius or cooldown_time timeout reaches zero
|
|
# (whichever happens first). Default: 40.
|
|
cooldown_temp: <temperature-threshold-for-cooldown>
|
|
# Timeout specified in minutes for CPU cooling down
|
|
# to cooldown_temp temperature. Zero value disables cooldown.
|
|
# Default: 0.
|
|
cooldown_time: <time-to-cooldown-in-minutes>
|
|
# CPU governor.
|
|
# See: https://www.kernel.org/doc/Documentation/cpu-freq/governors.txt
|
|
# for available values (they might differ for ARM and Intel).
|
|
governor: <one-of-scaling_available_governors-values>
|
|
# Restrict CPU usage to predefined "models":
|
|
# all, big_only, little_only, exclusive_cores.
|
|
cpu_usage: <usage-model>
|
|
# Setup CPU frequency as percent of max_freq.
|
|
# Default: 100
|
|
cpu_freq_pct: <0-100>
|
|
|
|
# The example below will run Telemetry toolchain performance benchmarks.
|
|
# The exact list of benchmarks that will be run can be seen in
|
|
# crosperf/experiment_factory.py
|
|
benchmark: all_toolchain_perf {
|
|
suite: telemetry_Crosperf
|
|
run_local: False
|
|
iterations: 1
|
|
}
|
|
|
|
# NOTE: You must specify at least one image; you may specify more than one.
|
|
# Replace <path-to-your-chroot-goes-here> and <board-goes-here> below.
|
|
vanilla_image {
|
|
chromeos_image:<path-to-your-chroot>/src/build/images/<board>/vanilla-image/chromiumos_test_image.bin
|
|
}
|
|
|
|
# Replace the chromeos image below with the actual path to your test image.
|
|
test_image {
|
|
chromeos_image:<path-to-your-chroot>/src/build/images/<board>/test-image/chromiumos_test_image.bin
|
|
}
|