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
Inline small but frequently invoked functions in PODArrayDetails #61144
Conversation
This is an automated comment for commit 7c44505 with description of existing statuses. It's updated for the latest CI running ❌ Click here to open a full report in a separate page
Successful checks
|
Let's wait for performance report in CI. |
(Seems a bit like microoptimization but why not. Can merge next week, CI is currently too broken.) |
I doubt the 2 slower perf tests are not related to inline of PODArrayDetails. I can't even find |
Yes. Perf tests seem not stable. I merged master again, and we now have different slower and faster cases. |
The merge button is greyed out for me because I could not merge from master as I don't have permissions to push to your repository:
Kindly merge from master by yourself, thanks! |
The RISC-V build is still broken, a fix is #61604. |
It is merged now. Let us see the newest CI |
@rschu1ze can we merge it now? Thanks! |
Before opt:
After opt:
Below graph is the flamegraph before optimization. It is obvious that the overhead of calling
byte_size
should be not ignored becausebyte_size
is on the hot path.Changelog category (leave one):
Changelog entry (a user-readable short description of the changes that goes to CHANGELOG.md):
More frequently invoked functions in PODArray are now force-inlined.