1794807 - 7.89 - 4.36% facebook-nav.landing ContentfulSpeedIndex / facebook-nav.groups ContentfulSpeedIndex + 1 more (Linux) regression on Sat October 8 2022

Status: RESOLVED WONTFIX

(Core :: Graphics: Text, defect)

Description

[Alex Finder]

Perfherder has detected a browsertime performance regression from push f1accadf1754ad5aa8d668ecb22ec8ef1a5bd880. 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)
8%	facebook-nav.landing ContentfulSpeedIndex	linux1804-64-shippable-qr	cold fission webrender	740.44 -> 798.83
5%	facebook-nav.groups ContentfulSpeedIndex	linux1804-64-shippable-qr	cold fission webrender	698.09 -> 731.62
4%	facebook-nav.groups ContentfulSpeedIndex	linux1804-64-shippable-qr	cold fission webrender	695.98 -> 726.33

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
7%	facebook-nav.landing ContentfulSpeedIndex	linux1804-64-shippable-qr	cold fission webrender	784.71 -> 726.65

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) may be backed out in accordance with our regression policy.

If you need the profiling jobs you can trigger them yourself from treeherder job view or ask a sheriff to do that for you.

For more information on performance sheriffing please see our FAQ.

Comment 1

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1791780

Comment 2

[Timothy Nikkel (:tnikkel)]

I would see if this regression still exists with bug 1735398 landed.

Comment 3

[Jonathan Kew [:jfkthame]]

I'm a bit confused by the report here -- what does it mean that we appear to have both a regression and an improvement for the same test (facebook-nav.landing / ContentfulSpeedIndex / linux1804-64-shippable-qr / cold fission webrender)?

Comment 4

[Timothy Nikkel (:tnikkel)]

Oh yeah. Good point. And two regressions in the same test. It looks like regressions/improvements that got assigned to a different alert were merged into one alert. Looking at the graph it is a bit of a mess around the time of the regression, but there is an overall regression that happened at some point.

Comment 5

[Ashley Hale [:ahale]]

Triage - tentatively marking this S3, it's not clear to me if this bug represents a regression or not, and whether the amount of regression justifies an S2 rating.

Comment 6

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1791780

Comment 7

[Alex Finder]

(In reply to Ashley Hale [:ahale] from comment #5)

Triage - tentatively marking this S3, it's not clear to me if this bug represents a regression or not, and whether the amount of regression justifies an S2 rating.

Hi Ashley! Are there any updates on this defect ?

Comment 8

[Ashley Hale [:ahale]]

I have basically no context on this bug, I was merely doing triage at the time. While the bug says this is a performance regression and will be backed out, I see no sign of backout when it landed a year ago, so I assume we simply accepted the performance loss from this bug. Is there any action we should be taking? Did the regressing bug impact performance of anything other than the COLR fonts that were previously not supported?

Comment 9

[Alex Finder]

The original graph doesn't show the original regression anymore, since it happened more than a year ago. And the original alert (accessible from the original comment on this bug) seems to have been deleted. Alex, how do you suggest we resolve the bug ? Should we close it as WONTFIX ?

Comment 10

[Alexandru Ionescu (needinfo me) [:alexandrui]]

There's nothing we can do if we can not see the data.