1462545 - 3.18% displaylist_mutate (linux64) regression on push e92d5e5095310ea59e3f787825732722b974c6f7 (Thu May 17 2018)

Status: RESOLVED WONTFIX

(Core :: Web Painting, defect)

Description

[Ionuț Goldan [:igoldan]]

Talos has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?changeset=e92d5e5095310ea59e3f787825732722b974c6f7

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

Regressions:

  3%  displaylist_mutate linux64 pgo e10s stylo     2,935.51 -> 3,028.73


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=13264

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]]

:peterv Perf regressions like the one above affected Windows as well. Can you please address this issue?

Comment 2

[Ionuț Goldan [:igoldan]]

Here are the Gecko profiles:

right before bug 1457301: https://perf-html.io/from-url/https%3A%2F%2Fqueue.taskcluster.net%2Fv1%2Ftask%2FDEQMnK9cRb2Ly9tGlg_joA%2Fruns%2F0%2Fartifacts%2Fpublic%2Ftest_info%2Fprofile_displaylist_mutate.zip

bug 1457301: https://perf-html.io/from-url/https%3A%2F%2Fqueue.taskcluster.net%2Fv1%2Ftask%2FIEIoFbzOQLiBj-rs3jpfOw%2Fruns%2F0%2Fartifacts%2Fpublic%2Ftest_info%2Fprofile_displaylist_mutate.zip

Comment 3

[Peter Van der Beken [:peterv]]

There's nothing actionable here. The patch in bug 1457301 is restoring some checks that were accidentally removed by the patches in bug 888600. If those are a performance issue they were before bug 888600 landed too.

Comment 4

[Ionuț Goldan [:igoldan]]

(In reply to Peter Van der Beken [:peterv] from comment #3)
> There's nothing actionable here. The patch in bug 1457301 is restoring some
> checks that were accidentally removed by the patches in bug 888600. If those
> are a performance issue they were before bug 888600 landed too.

I didn't see any perf changes before or after bug 888600 landed in [1]. So from the figures, I wouldn't say we lived with them so far.


[1] https://bugzilla.mozilla.org/show_bug.cgi?id=888600#c99

Comment 5

[Andrew McCreight [:mccr8]]

Matt, do you have any idea why adding a check to message manager would affect displaylist_mutate? Off hand, it doesn't feel very related. Thanks.

I'll move this to a layout component, so the people who might care about this are more likely to see it.

Comment 6

[Matt Woodrow (:mattwoodrow)]

I think the range is wrong here. The previous push to inbound [1] seems much more likely, especially https://hg.mozilla.org/integration/mozilla-inbound/rev/488b7be0348b


[1] https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=97ca42a0634b14ad9519612f4cad822de6dc26bf&tochange=2a647313f49008f277739cb7fce6209bff235d03

Comment 7

[Matt Woodrow (:mattwoodrow)]

That patch landed some debugging code (that won't be compiled in for beta/release), so I think we should accept the regression for Nightly.

Comment 8

[Ionuț Goldan [:igoldan]]

(In reply to Matt Woodrow (:mattwoodrow) from comment #6)
> I think the range is wrong here. The previous push to inbound [1] seems much
> more likely, especially
> https://hg.mozilla.org/integration/mozilla-inbound/rev/488b7be0348b
> 
> 
> [1]
> https://hg.mozilla.org/integration/mozilla-inbound/
> pushloghtml?fromchange=97ca42a0634b14ad9519612f4cad822de6dc26bf&tochange=2a64
> 7313f49008f277739cb7fce6209bff235d03

I think I agree. I'll do some retriggers to properly confirm this.

Comment 9

[Mike Taylor [:miketaylr]]

(In reply to Ionuț Goldan [:igoldan], Performance Sheriffing from comment #8)
> (In reply to Matt Woodrow (:mattwoodrow) from comment #6)
> > I think the range is wrong here. The previous push to inbound [1] seems much
> > more likely, especially
> > https://hg.mozilla.org/integration/mozilla-inbound/rev/488b7be0348b
> > 
> > 
> > [1]
> > https://hg.mozilla.org/integration/mozilla-inbound/
> > pushloghtml?fromchange=97ca42a0634b14ad9519612f4cad822de6dc26bf&tochange=2a64
> > 7313f49008f277739cb7fce6209bff235d03
> 
> I think I agree. I'll do some retriggers to properly confirm this.

Any updates?

Comment 10

[Ionuț Goldan [:igoldan]]

(In reply to Mike Taylor [:miketaylr] (62 Regression Engineering Owner) from comment #9)
> (In reply to Ionuț Goldan [:igoldan], Performance Sheriffing from comment #8)
> > (In reply to Matt Woodrow (:mattwoodrow) from comment #6)
> > > I think the range is wrong here. The previous push to inbound [1] seems much
> > > more likely, especially
> > > https://hg.mozilla.org/integration/mozilla-inbound/rev/488b7be0348b
> > > 
> > > 
> > > [1]
> > > https://hg.mozilla.org/integration/mozilla-inbound/
> > > pushloghtml?fromchange=97ca42a0634b14ad9519612f4cad822de6dc26bf&tochange=2a64
> > > 7313f49008f277739cb7fce6209bff235d03
> > 
> > I think I agree. I'll do some retriggers to properly confirm this.
> 
> Any updates?

:mattwood got it right in comment 6.