1476265 - 11.71 - 11.97% installer size (windows2012-32, windows2012-64) regression on push e09ba0bb848afe8b66c4cdd56794301b790c6c34 (Tue Jul 17 2018)

Status: RESOLVED WONTFIX

(Firefox Build System :: General, defect)

Description

[Ionuț Goldan [:igoldan]]

We have detected a build metrics regression from push:

https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?changeset=e09ba0bb848afe8b66c4cdd56794301b790c6c34

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

Regressions:

 12%  installer size windows2012-64 pgo      63,552,057.92 -> 71,162,374.25
 12%  installer size windows2012-32 pgo      60,381,909.83 -> 67,452,763.08


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=14376

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

[(Away)]

Much like bug 1474860, this is another expected increase in size in exchange for faster speed. I'm looking forward to seeing the Talos results.

Comment 2

[Nathan Froyd [:froydnj]]

10%+ for any other reason would be backed out ASAP, so this is going to need some significant justification to stay in the tree.

Comment 3

[Anthony Jones (:ajones, :kentuckyfriedtakahe, :k17e)]

(In reply to Nathan Froyd [:froydnj] from comment #2)
> 10%+ for any other reason would be backed out ASAP, so this is going to need
> some significant justification to stay in the tree.

I don't think we can consider this separate from 1474860 because it fixes the perf regressions. The current plan is to assess the full picture when we've got PGO suppport.

Comment 4

[Ionuț Goldan [:igoldan]]

(In reply to Anthony Jones (:kentuckyfriedtakahe, :k17e) from comment #3)
> (In reply to Nathan Froyd [:froydnj] from comment #2)
> > 10%+ for any other reason would be backed out ASAP, so this is going to need
> > some significant justification to stay in the tree.
> 
> I don't think we can consider this separate from 1474860 because it fixes
> the perf regressions. The current plan is to assess the full picture when
> we've got PGO suppport.

Can you give an estimation for when we'll be able to assess the full picture? How much time until then?

Comment 5

[(Away)]

Enabling PGO in bug 1341525 partially mitigated this regression: we won back 2.8MB/4% of installer size. I don't know why Perfherder didn't raise an alert to this.

Comment 6

[Ionuț Goldan [:igoldan]]

(In reply to David Major [:dmajor] from comment #5)
> Enabling PGO in bug 1341525 partially mitigated this regression: we won back
> 2.8MB/4% of installer size. I don't know why Perfherder didn't raise an
> alert to this.

We now have PGO support. Is there ongoing work concerning the installer sizes?

Comment 7

[Sylvestre Ledru [:Sylvestre]]

Just like in bug 1474860
I think our users prefer to have a much faster Firefox than a smaller Firefox.
I think we should accept this trade-off.