1909258 - 18.52 - 18.32% AndroidStartup:fenix cold_main_first_frame.median / AndroidStartup:fenix cold_main_first_frame.mean (Android) regression on Wed July 10 2024

Status: NEW

(Fenix :: Toolbar, defect, P1)

Description

[Andrej (:aglavic)]

Perfherder has detected a mozperftest performance regression from push 2ed46bf1ae1d87e34ff254b919da6e1f607f25de. 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)
19%	AndroidStartup:fenix cold_main_first_frame.median	android-hw-a51-11-0-aarch64-shippable		979.00 -> 1,160.29
18%	AndroidStartup:fenix cold_main_first_frame.mean	android-hw-a51-11-0-aarch64-shippable		990.59 -> 1,172.08

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 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.

You can run these tests on try with ./mach try perf --alert 1185

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 1902798

Comment 2

[Chris Peterson [:cpeterson]]

S2 because this 20% startup regression affects all Nightly users.

Comment 3

[Joe M [:jmahon]]

Roger, for the sake of the 7/30 sprint, can you summarize - as precisely as possible - what you expect to be able to accomplish by 8/6? (e.g. what you will have investigated; even if we haven't resolved the issue itself)

Comment 4

[Roger Yang [:royang]]

From the performance profiles, there are three things that requires more investigation:

1. Felt privacy Nimbus feature check and exposure is costing some performance regression. I will work with the product team to determine if we can end that experiment and ship felt privacy without Nimbus.
2. Long press button written in compose seems to take longer to inflate, I will investigate to see if this is an issue or a red herring.
3. Tab strip check seems expensive, but I'll have to determine if that is something we'll have to look into for navigation bar performance regression.

Comment 5

[Roger Yang [:royang]]

https://share.firefox.dev/3Sp0Jfi with navigation bar
https://share.firefox.dev/3YgW6rr without navigation bar

Comment 6

[Joe M [:jmahon]]

Note about our plan for handling this regression: https://docs.google.com/document/d/1qWPNmjldbubYrUlA8PtuET_ANYlusTskXGWlBL6Dk0Q/edit

(long story short: we're leaving the regression on Nightly for now but treating resolving it as our top priority)

Comment 7

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

Set release status flags based on info from the regressing bug 1902798