1461942 - 92.03 - 199.6% tp5o_scroll / tscrollx (linux64-qr, windows10-64-qr) regression on push 90445f7b62cdb1deda8e3a860032c9f113de3262 (Tue May 15 2018)

Status: RESOLVED WONTFIX

(Core :: Graphics: WebRender, defect)

Description

[Ionuț Goldan [:igoldan]]

Talos has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=90445f7b62cdb1deda8e3a860032c9f113de3262

As author of one of the patches included in that push, we need your help to address this regression.

Regressions:

200%  tp5o_scroll linux64-qr opt e10s stylo     0.18 -> 0.54
186%  tp5o_scroll windows10-64-qr opt e10s stylo0.16 -> 0.46
 92%  tscrollx linux64-qr opt e10s stylo        0.22 -> 0.42


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=13230

On the page above you can see an alert for each affected platform as well as a link to a graph showing the history of scores for this test. There is also a link to a treeherder page showing the Talos jobs in a pushlog format.

To learn more about the regressing test(s), please see: https://wiki.mozilla.org/Buildbot/Talos/Tests

For information on reproducing and debugging the regression, either on try or locally, see: https://wiki.mozilla.org/Buildbot/Talos/Running

*** Please let us know your plans within 3 business days, or the offending patch(es) will be backed out! ***

Our wiki page outlines the common responses and expectations: https://wiki.mozilla.org/Buildbot/Talos/RegressionBugsHandling

Comment 1

[Ionuț Goldan [:igoldan]]

:kats Bug 1461122 caused the perf regressions you see above. Can we fix them or should we consider accepting or backing them out?

Note: I am not really sure about the seriousness of the regressions. The percentages do look worrisome, still the absolute values are quite small.

Comment 2

[Kartikaya Gupta (email:[e-mail])]

We have to accept the perf regression as it fixes a correctness issue in the original patch (bug 1452390) which had a massive perf win. Note that the new numbers even after this regression are better than the old numbers before the original patch (see bug 1452390 comment 27).

Closing as WONTFIX.