Talos has detected a Firefox performance regression from push 729307f6ac594f76d1de7d7697ca5c3880135de3. As author of one of the patches included in that push, we need your help to address this regression. Regressions: 17% tp5n main_normal_fileio windows7-32 opt 53122011.5 -> 62027064.67 17% tart summary windowsxp opt 7.1 -> 8.29 15% tscrollx summary windowsxp opt 2.88 -> 3.32 15% tsvgr_opacity summary windows7-32 opt 696.28 -> 797.24 14% tsvgr_opacity summary windowsxp opt 686.7 -> 785.06 11% tp5o_scroll summary windowsxp opt 4.33 -> 4.81 6% tresize windowsxp opt e10s 12.57 -> 13.28 4% tp5o Private Bytes windows7-32 opt 196049105.48 -> 204616966.67 3% tp5o Main_RSS windows7-32 opt 164384440.84 -> 169192925.22 2% cart summary windowsxp opt e10s 38.56 -> 39.51 2% cart summary windows7-32 opt 36.81 -> 37.67 2% tart summary windowsxp opt e10s 7.7 -> 7.87 Improvements: 38% tsvgr_opacity summary windowsxp opt e10s 1054.2 -> 654.71 31% tsvgx summary windowsxp opt e10s 274.86 -> 190.3 26% tsvgx summary windows7-32 opt 462.8 -> 343.8 26% tsvgx summary windowsxp opt 473.03 -> 351.86 18% tps summary windowsxp opt e10s 48.34 -> 39.43 15% cart summary windowsxp opt 46.11 -> 39.42 10% tps summary windows7-32 opt 78.53 -> 71.02 7% tps summary windowsxp opt 73.1 -> 67.84 You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=3363 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

After doing some retriggers, this issue might be caused by https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=717d85fc2046149b2a3443ffcc7aff04e119b0f7&tochange=729307f6ac594f76d1de7d7697ca5c3880135de3 Hi Andrew, as you are the patch author, can you take a look at this and determine what is the root cause? Thanks!

this looks identical to what mchang did a couple weeks ago, not sure if there are any lessons learned there (bug 1298608). He indicated he would reland in mid october- this is still september so either things went well, or this is something different.

(In reply to Joel Maher ( :jmaher) from comment #2) > this looks identical to what mchang did a couple weeks ago, not sure if > there are any lessons learned there (bug 1298608). He indicated he would > reland in mid october- this is still september so either things went well, > or this is something different. This is the skia telemetry experiment. From bug 1298608, I think we're going to have to take the regressions as we couldn't see any particular reason for the font file read differences.

ignoring windows xp and the file access, I see: 15% tsvgr_opacity summary windows7-32 opt that seems like a large regression- is that of concern?

(In reply to Joel Maher ( :jmaher) from comment #4) > ignoring windows xp and the file access, I see: > 15% tsvgr_opacity summary windows7-32 opt > > that seems like a large regression- is that of concern? I'm less worried about it because from bug 1298608, tsvgr_opacity with e10s is a 39% improvement. Since this won't ship until 52, iirc, we will want e10s everywhere by then.

Sounds like this might be a wontfix?

I will let Mason make the call on wontfix- but it does sound like a wontfix.