5.44% reddit-billgates-post-2.top LastVisualChange (Windows) regression on Tue January 25 2022
Categories
(Core :: JavaScript: GC, defect)
Tracking
()
| Tracking | Status | |
|---|---|---|
| firefox98 | --- | affected |
People
(Reporter: afinder, Unassigned)
References
(Regression)
Details
(Keywords: perf, perf-alert, regression)
Perfherder has detected a browsertime performance regression from push 2c00bc26b97a10e0767b614e4ba89047941dcc18. As author of one of the patches included in that push, we need your help to address this regression.
Regressions:
| Ratio | Test | Platform | Options | Absolute values (old vs new) |
|---|---|---|---|---|
| 5% | reddit-billgates-post-2.top LastVisualChange | windows10-64-shippable-qr | cold fission webrender | 737.56 -> 777.71 |
Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests. Please follow our guide to handling regression bugs and let us know your plans within 3 business days, or the offending patch(es) will be backed out in accordance with our regression policy.
For more information on performance sheriffing please see our FAQ.
|
||
Updated•2 years ago
|
|
||
Comment 1•2 years ago
|
||
Bug 1751162 aims to improve responsiveness by reducing the number of long minor GCs, and the latest telemetry shows that it is working. When making changes to the heuristics that govern GC there are always a few performance changes in both directions. Given that this change is effective and that there have been only two small regressions, I'd like to accept this.
Description
•