-
Notifications
You must be signed in to change notification settings - Fork 29.7k
const vs. non-const widget build benchmark #148261
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
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why is markNeedsBuild also timed?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That's what we do in our other build benchmarks, which I used for a template for this one.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: this can be a bit spammy, maybe add this to printer?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ups, I just forgot to remove this after doing some debugging.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
this benchmark running on a Moto G4 shows that const'ness is 5-7% faster
Rebuilding 30+ elements, plus overhead is only 5-7% slower than rebuilding 1 element, plus overhead, so it looks like the majority of the cpu time is spent on the overhead? Given that there's no relayout or repaint maybe instead of pumpBenchmark we should run owner.buildScope to increase the signal to noise ratio?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Let's see what the benchmark numbers will look like in the lab and we can then tweak this if it is too noisy.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hmm if an operation used to take 10s but now takes 5s I thought that should be called a 100% speed up? But with this formula that would only be a 50% speedup?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Oh, right. I'll fix this up.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done. Can you check my updated math?
LongCatIsLooong
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
still LGTM
76e0814 to
b177353
Compare
flutter/flutter@0d22d91...00425ef 2024-05-17 [email protected] Roll Flutter Engine from d575e511f9e5 to a19d3722922d (2 revisions) (flutter/flutter#148534) 2024-05-17 [email protected] Roll Flutter Engine from 8d1a1d8d7b48 to d575e511f9e5 (5 revisions) (flutter/flutter#148529) 2024-05-17 [email protected] Manual roll Flutter Engine from 6fa734d68688 to 8d1a1d8d7b48 (8 revisions) (flutter/flutter#148528) 2024-05-17 [email protected] Add test for material_banner.0.dart and material_banner.1.dart (flutter/flutter#148452) 2024-05-17 [email protected] `switch` statement cleanup (flutter/flutter#148382) 2024-05-16 [email protected] const vs. non-const widget build benchmark (flutter/flutter#148261) 2024-05-16 49699333+dependabot[bot]@users.noreply.github.com Bump actions/checkout from 4.1.5 to 4.1.6 (flutter/flutter#148516) 2024-05-16 98614782+auto-submit[bot]@users.noreply.github.com Reverts "Roll Flutter Engine from 6fa734d68688 to 1850def2ccff (1 revision) (#148507)" (flutter/flutter#148517) 2024-05-16 [email protected] Add PrivacyInfo.xcprivacy to plugin template (flutter/flutter#148485) 2024-05-16 [email protected] Fix iOS reference in macOS Cocoapods error (flutter/flutter#148506) 2024-05-16 [email protected] Roll Flutter Engine from 6fa734d68688 to 1850def2ccff (1 revision) (flutter/flutter#148507) 2024-05-16 [email protected] Roll Flutter Engine from 460df6caef0e to 6fa734d68688 (4 revisions) (flutter/flutter#148500) 2024-05-16 [email protected] Enhanced enum features for `AnimationStatus` (flutter/flutter#147801) 2024-05-16 [email protected] [macOS] codesign native assets during embed (flutter/flutter#148310) If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/flutter-packages Please CC [email protected],[email protected] on the revert to ensure that a human is aware of the problem. To file a bug in Packages: https://github.com/flutter/flutter/issues/new/choose To report a problem with the AutoRoller itself, please file a bug: https://issues.skia.org/issues/new?component=1389291&template=1850622 Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md
flutter/flutter@0d22d91...00425ef 2024-05-17 [email protected] Roll Flutter Engine from d575e511f9e5 to a19d3722922d (2 revisions) (flutter/flutter#148534) 2024-05-17 [email protected] Roll Flutter Engine from 8d1a1d8d7b48 to d575e511f9e5 (5 revisions) (flutter/flutter#148529) 2024-05-17 [email protected] Manual roll Flutter Engine from 6fa734d68688 to 8d1a1d8d7b48 (8 revisions) (flutter/flutter#148528) 2024-05-17 [email protected] Add test for material_banner.0.dart and material_banner.1.dart (flutter/flutter#148452) 2024-05-17 [email protected] `switch` statement cleanup (flutter/flutter#148382) 2024-05-16 [email protected] const vs. non-const widget build benchmark (flutter/flutter#148261) 2024-05-16 49699333+dependabot[bot]@users.noreply.github.com Bump actions/checkout from 4.1.5 to 4.1.6 (flutter/flutter#148516) 2024-05-16 98614782+auto-submit[bot]@users.noreply.github.com Reverts "Roll Flutter Engine from 6fa734d68688 to 1850def2ccff (1 revision) (#148507)" (flutter/flutter#148517) 2024-05-16 [email protected] Add PrivacyInfo.xcprivacy to plugin template (flutter/flutter#148485) 2024-05-16 [email protected] Fix iOS reference in macOS Cocoapods error (flutter/flutter#148506) 2024-05-16 [email protected] Roll Flutter Engine from 6fa734d68688 to 1850def2ccff (1 revision) (flutter/flutter#148507) 2024-05-16 [email protected] Roll Flutter Engine from 460df6caef0e to 6fa734d68688 (4 revisions) (flutter/flutter#148500) 2024-05-16 [email protected] Enhanced enum features for `AnimationStatus` (flutter/flutter#147801) 2024-05-16 [email protected] [macOS] codesign native assets during embed (flutter/flutter#148310) If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/flutter-packages Please CC [email protected],[email protected] on the revert to ensure that a human is aware of the problem. To file a bug in Packages: https://github.com/flutter/flutter/issues/new/choose To report a problem with the AutoRoller itself, please file a bug: https://issues.skia.org/issues/new?component=1389291&template=1850622 Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md
|
Reason for revert: #148845 |
|
Time to revert pull request flutter/flutter/148261 has elapsed. |
This reverts commit c4ad1dc.
|
This does not imply any significant impact on a real app. It shows something, just not sure that this something is relevant to flutter users. |
Reverts flutter#148261 For some reason, the benchmark is flaky only on iOS. Android runs are fine. See flutter#148845. Closes flutter#148845.
This benchmark illustrates the impact that const'ness has on widget build times by building a widget tree with all const-widgets and then building the same widget tree without any const'ness. Local testing with this benchmark running on a Moto G4 shows that const'ness is 13-16% faster.
I'd like to check this benchmark in because the question of how important widget const'ness is comes up every so often. With this benchmark we have up-to-date data to point people to.