6.28 - 5.16% buzzfeed FirstVisualChange / buzzfeed largestContentfulPaint (Linux) regression on Mon April 28 2025
Categories
(Core :: CSS Parsing and Computation, defect)
Tracking
()
People
(Reporter: intermittent-bug-filer, Unassigned)
References
(Regression)
Details
(Keywords: perf, perf-alert, regression)
Perfherder has detected a browsertime performance regression from push 8042121f2dcefb7699d518ff2909c54682c88cb0. As author of one of the patches included in that push, we need your help to address this regression.
Please acknowledge, and begin investigating this alert within 3 business days, or the patch(es) may be backed out in accordance with our regression policy. Our guide to handling regression bugs has information about how you can proceed with this investigation.
If you have any questions or need any help with the investigation, please reach out to aesanu@mozilla.com. Alternatively, you can find help on Slack by joining #perf-help, and on Matrix you can find help by joining #perftest.
Regressions:
| Ratio | Test | Platform | Options | Absolute values (old vs new) | Performance Profiles |
|---|---|---|---|---|---|
| 6% | buzzfeed FirstVisualChange | linux1804-64-shippable-qr | fission warm webrender | 152.90 -> 162.51 | |
| 5% | buzzfeed largestContentfulPaint | linux1804-64-shippable-qr | fission warm webrender | 127.74 -> 134.33 |
Improvements:
| Ratio | Test | Platform | Options | Absolute values (old vs new) | Performance Profiles |
|---|---|---|---|---|---|
| 14% | bing-search loadtime | macosx1470-64-shippable | bytecode-cached fission warm webrender | 77.43 -> 66.46 | Before/After |
| 12% | bing-search largestContentfulPaint | macosx1470-64-shippable | bytecode-cached fission warm webrender | 97.47 -> 85.63 | Before/After |
| 10% | imdb fcp | linux1804-64-shippable-qr | cold fission webrender | 742.16 -> 670.45 | Before/After |
| 8% | ebay fcp | linux1804-64-shippable-qr | fission warm webrender | 54.37 -> 50.27 | Before/After |
| 7% | ebay fcp | linux1804-64-shippable-qr | fission warm webrender | 53.46 -> 49.64 | Before/After |
| ... | ... | ... | ... | ... | ... |
| 2% | speedometer3 Editor-CodeMirror/Long/total | macosx1500-aarch64-shippable | fission webrender | 7.54 -> 7.37 | Before/After |
Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests.
If you need the profiling jobs you can trigger them yourself from treeherder job view or ask aesanu@mozilla.com to do that for you.
You can run all of these tests on try with ./mach try perf --alert 44967
The following documentation link provides more information about this command.
|
||
Comment 1•9 months ago
|
||
Are there profiles for the regression? But anyways given this is overwhelmingly an improvement I think we should keep it.
|
||
Comment 2•8 months ago
|
||
Hello. This is the performance Firefox Profiler for the Buzzfeed regression.
This is the profiler job link for Buzzfeed.
This is the link for the previous profile job, push id 6fb1d0ccf4393763f55542c9acde2f85e2a287b6
|
|
||
Comment 3•8 months ago
|
||
Interesting. So that seems to have:
<style><!-- massive style --></style>
<style><!-- massive style --></style>
<link rel="preload" href="/static-assets/css/feeds.1cc6b001b92c13c51684.css?brotli=allow" as="style" onload="this.rel='stylesheet'" onerror="window.bfa('track/website/instrumentation', { data: { target: 'stylesheet', value: 'timeout' } })">
And the load event causes an invalidation that isn't on the "before" page, which is ~10ms which explains the regression...
So I think this is about the timing the preload link makes it into the DOM from the parser, and given that I think this is not really actionable on the style end. That combined with the multiple improvements in here and bug 1956486 comment 5, I think we can safely WONTFIX the buzzfeed regression.
Description
•