-
-
Notifications
You must be signed in to change notification settings - Fork 1.2k
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
Set up a benchmark for our trampolines. #519
Conversation
This commit adds a new benchmark to the bench project. It updates the version of JMH that we are using, and also adds a new, hidden Eval subtype which improves the performance of Eval.defer.
Here's the benchmark result running on my machine:
(Run with |
It seems like TailCalls in 2.10 lacks .flatMap and .map, meaning that our benchmark won't work. If we comment it out for now, we can eventually use cats.macros.Platform to compile this benchmark only on 2.11.
Are the build failures random or related to this? |
I'm looking into it. It doesn't seem like Travis flakiness. |
So |
I restarted the builds and this time they worked in 2.10 but not 2.11. It does seem like a quirky sporadic/resource issue, but if we aren't seeing this outside of this branch, I'm also a little hesitant to merge this PR simply because I don't want us to start completely ignoring build statuses. @inthenow do you have any ideas? |
Maybe we should set up Contained-Based builds? I wonder if that would help: http://blog.travis-ci.com/2014-12-17-faster-builds-with-container-based-infrastructure/ EDIT: This article looks old -- maybe we are already doing this. I need to investigate further. |
Current coverage is
|
Current coverage is
|
Awesome. We seem to have overcome the previous build flakiness! |
👍 |
1 similar comment
👍 |
Set up a benchmark for our trampolines.
This commit adds a new benchmark to the bench project.
It updates the version of JMH that we are using, and
also adds a new, hidden Eval subtype which improves
the performance of Eval.defer.