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

Program configuration optimizations #3489

Merged
merged 3 commits into from
Oct 30, 2016
Merged

Program configuration optimizations #3489

merged 3 commits into from
Oct 30, 2016

Conversation

mourner
Copy link
Member

@mourner mourner commented Oct 30, 2016

Along with #3485, this PR brings the rendering performance with the Streets z13 benchmark to pre-DDS levels and closes #3342! 🎉🎉🎉 Benchmark results:

Changes:

  • Avoid repeatedly restringifying program configurations for caching.
  • Avoid repeatedly recreating the same fill program for clipping masks and background.

Note that the failing build is due to changes in ESLint 3.9.0; fixed up the warnings in the third commit.

👀 @jfirebaugh @lucaswoj @ansis

@mourner mourner added the performance ⚡ Speed, stability, CPU usage, memory usage, or power usage label Oct 30, 2016
Copy link
Contributor

@jfirebaugh jfirebaugh left a comment

Choose a reason for hiding this comment

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

Awesome!

@mourner
Copy link
Member Author

mourner commented Oct 30, 2016

@jfirebaugh oops, nope, the two render tests for switching paint from default to property function are failing. Investigating.

@mourner mourner force-pushed the program-optimizations branch from e9613dd to d47dfcc Compare October 30, 2016 02:42
@mourner
Copy link
Member Author

mourner commented Oct 30, 2016

OK fixed the bug and rebased, merging on green.

@mourner mourner merged commit ee13651 into master Oct 30, 2016
@mourner mourner deleted the program-optimizations branch October 30, 2016 02:52
@mourner
Copy link
Member Author

mourner commented Oct 30, 2016

For the record, frame-duration benchmark results:

before both PRs: 9.6 ms, 5% > 16ms
after: 7.4 ms, 0% > 16ms

@mourner
Copy link
Member Author

mourner commented Oct 30, 2016

Note that all these optimizations became much more obvious after the ProgramConfiguration #3439 refactor. It made understanding how programs are created and used much easier.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
performance ⚡ Speed, stability, CPU usage, memory usage, or power usage
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Address rendering performance regressions after DDS
2 participants