1453081 - 5.24% tpaint (osx-10-10) regression on push 1fdb4dcb4225228e5ebc36c3de4d259fe730aedf (Mon Apr 9 2018)

Status: RESOLVED WONTFIX

(Firefox :: Tours, defect)

Description

[Joel Maher ( :jmaher ) (UTC -8)]

Talos has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=1fdb4dcb4225228e5ebc36c3de4d259fe730aedf

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

Regressions:

  5%  tpaint osx-10-10 opt e10s stylo     304.29 -> 320.23


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=12610

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

[Joel Maher ( :jmaher ) (UTC -8)]

:kmag, I see you authored the patch for bug 1440747.  That seems to be the root cause of this talos regression.  Is this expected?  Is there something you could do here to reduce or fix the regression?  Should we accept this regression?

Comment 2

[Joel Maher ( :jmaher ) (UTC -8)]

for reference, this is only seen on osx- and for osx the test was sort of bi-modal and now we are posting data on the higher mode.  So good news is the test is not as noisy as before!

Comment 3

[Kris Maglione [:kmag]]

I think this is probably just noise that we should accept. The over-all impact of this change should be better performance and memory usage. There might be a very small amount of additional overhead the first time about:newtab or about:home is loaded in a new process, but there will also be considerably less overhead whenever any other page is loaded. For OS-X, it looks like that slight timing change was just enough to make it behave more consistently.

Comment 4

[Ryan VanderMeulen [:RyanVM]]

Comment 3 sounds plausible, and indeed the numbers do appear to be more stable now around what used to be the upper end of the bimodality range. Joel and I both agree that this doesn't look worth spending more time investigating based on that.