Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

lisa.tests.cpufreq.schedutil: Add LargeStepUp test for ramp boost fea… #1011

Merged

Conversation

douglas-raillard-arm
Copy link
Contributor

@douglas-raillard-arm douglas-raillard-arm commented Aug 5, 2019

…ture

Provide a test for the ramp boost feature as posted on LKML.
This test will mostly catch cases where schedutil is not called as often as
expected, since it reads most of the signals from the kernel instead of guessing
them.

lisa/tests/staging/schedutil.py Outdated Show resolved Hide resolved
lisa/tests/staging/schedutil.py Outdated Show resolved Hide resolved
lisa/tests/staging/schedutil.py Show resolved Hide resolved
lisa/tests/staging/schedutil.py Show resolved Hide resolved
lisa/tests/staging/schedutil.py Show resolved Hide resolved
# Tolerate a few bad samples that added too much boost
passed_shape = bad_shape_pct < bad_samples_threshold_pct

passed = passed_overhead and passed_shape
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Would it make sense to split these into two separate tests? We'd get finer grained results.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not sure if that would really help, all these checks are just checking the overall shape of the signal. If I could simulate the signals with 100% accuracy, they would all go away and replaced by a Series equality check.


# Ensure accurate duty cycle and idle state misprediction on some
# boards. This helps having predictable execution.
with target.disable_idle_states():
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We keep carrying over an explanation as to why we do this... Perhaps we should stick it in a load-tracking specific "I care about signals so give me a white room` context manager.

Copy link
Contributor Author

@douglas-raillard-arm douglas-raillard-arm Sep 27, 2019

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good cleanup idea, added to the TODO list

…ture

Provide a test for the ramp boost feature as posted on LKML.
This test will mostly catch cases where schedutil is not called as often as
expected, since it reads most of the signals from the kernel instead of guessing
them.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants