1371609 - 1.12 - 2.5% build times / compiler warnings (linux32, linux64) regression on push a344042d76e92e824f153da614d6571d63a2c71b (Fri Jun 9 2017)

Status: RESOLVED WONTFIX

(Core :: Audio/Video: Playback, defect, P3)

Description

[Ionuț Goldan [:igoldan]]

We have detected a build metrics regression from push:

https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=a344042d76e92e824f153da614d6571d63a2c71b

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

Regressions:

  3%  build times summary linux32 pgo taskcluster-c4.4xlarge     2,940.39 -> 3,014.01
  1%  compiler warnings summary linux64 asan asan debug          380.92 -> 385.17


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=7135

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://developer.mozilla.org/en-US/docs/Mozilla/Performance/Automated_Performance_Testing_and_Sheriffing/Build_Metrics

Comment 1

[Ralph Giles (:rillian)]

We added a new 3rd-party library, so I'm not surprised by the build time increase. (But only on linux32 pgo?)

Warnings we can fix.

Comment 2

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

Just saw an installer size alert show up for this revision:
== Change summary for alert #7135 (as of June 08 2017 21:27 UTC) ==

Regressions:

  3%  build times summary linux32 pgo taskcluster-c4.4xlarge     2,940.39 -> 3,014.01
  1%  installer size summary linux32 opt                         59,684,253.08 -> 60,392,248.42
  1%  compiler warnings summary linux64 asan asan debug          380.92 -> 385.17

For up to date results, see: https://treeherder.mozilla.org/perf.html#/alerts?id=7135

Letting :sylvestre know so he can track this.

Comment 3

[Sylvestre Ledru [:Sylvestre]]

Can we build/ship only the part that we use in aom? (if it is not the case)

Comment 4

[Ralph Giles (:rillian)]

We're currently using only the decoder, so I could try patching out the encoder until we need it for the WebRTC and MediaRecorder APIs. However, this is a nightly-only experiment; the regression should be resolved by the code turning off entirely when Firefox 55 goes to beta.

Comment 5

[Ralph Giles (:rillian)]

This change is nightly-only so it shouldn't affect Firefox 56 or 57 once it goes to beta.

Comment 6

[Ralph Giles (:rillian)]

Same story through Firefox 60; this is still nightly-only code.