1497785 - 4.54% Heap Unclassified (windows10-64) regression on push 72e7ef77480d0845ceabb5d3747a5523db6bda9d (Wed Oct 10 2018)

Status: RESOLVED WONTFIX

(Core :: DOM: Content Processes, defect)

Description

[Ionuț Goldan [:igoldan]]

We have detected an awsy regression from push:

https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?changeset=72e7ef77480d0845ceabb5d3747a5523db6bda9d

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

Regressions:

  5%  Heap Unclassified windows10-64 pgo stylo     37,429,954.61 -> 39,127,824.39

Improvements:

 12%  Images windows10-64 pgo stylo     7,936,538.05 -> 7,021,840.48


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=16599

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 jobs in a pushlog format.

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

Comment 1

[Ionuț Goldan [:igoldan]]

This also happens for Linux 64bit and Windows 7.

Comment 2

[Kris Maglione [:kmag]]

This is an expected side-effect of increasing the process count.

Comment 3

[Ionuț Goldan [:igoldan]]

This also brought tps regressions on some of the desktop platforms:

== Change summary for alert #16593 (as of Tue, 09 Oct 2018 18:07:51 GMT) ==

Regressions:

 19%  tps windows7-32 pgo e10s stylo     10.71 -> 12.77
 11%  tps windows7-32 opt e10s stylo     12.13 -> 13.49
  7%  tps linux64 opt e10s stylo         10.80 -> 11.61
  7%  tps linux64-qr opt e10s stylo      9.18 -> 9.80

Improvements:

 13%  tps windows10-64 pgo e10s stylo        11.72 -> 10.17
  9%  tps windows10-64 opt e10s stylo        12.44 -> 11.30
  7%  tps windows10-64-qr opt e10s stylo     9.58 -> 8.88

For up to date results, see: https://treeherder.mozilla.org/perf.html#/alerts?id=16593

Comment 4

[Ionuț Goldan [:igoldan]]

:erahm were these also expected?

Comment 5

[Eric Rahm [:erahm]]

mconley, bumping to 8 content processes seems to have both regressed *and* improved the talos tps test. Is this something we want to look into further?

Comment 6

[Mike Conley (:mconley) (:⚙️)]

Yeah, I think I'd like to understand this. I've filed bug 1501825 for that.