Talos has detected a Firefox performance regression from push: https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=f87e31ac8e9a3588e09a62bbb092806d75b324df As author of one of the patches included in that push, we need your help to address this regression. Regressions: 18% sessionrestore linux64 opt e10s 644.67 -> 759.83 16% sessionrestore_no_auto_restore linux64 opt e10s689.42 -> 799.25 14% sessionrestore linux64 pgo e10s 559.17 -> 639.42 14% sessionrestore_no_auto_restore linux64 pgo e10s596.83 -> 682.08 12% ts_paint linux64 opt e10s 1,055.00 -> 1,183.08 11% ts_paint linux64 pgo e10s 939.08 -> 1,038.58 You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=7580 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

this was backfilled and shows a clear root cause: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=851550a08fc118255228a72bda0f7755372698e0&tochange=7b6b179caf46386f2b40066b2dc3e3744bdcd6ba :milan, I see that you authored the patch which caused the regression- can you take a look at this?

I believe this is the equivalent of bug 1356138, from the patches on bug 1337063, but on Linux, vs. Windows. As we remove sync messages from content to parent on startup, it seems that we end up giving more work to the parent process (sync message is otherwise throttling the content process.) Between bug 1337703 and bug 1337062 we remove two such sync messages on startup - it seems the first one gets us to see the talos regression on Windows, while the second one gets us to see the talos regression on Linux. I do want to do a try run to confirm this; give me a couple of days to get that confirmed before backing this out.

Sounds good Milan, thanks for the note.

https://treeherder.mozilla.org/#/jobs?repo=try&revision=4c5d6a45b68fe42fc1bac95ec39111674ff61718

Looking at the talos run from comment 4, my suspicion seems to be correct. Adding back in a sync call improves the talos results back. So, this really is the same as "regression" from bug 1337063 tracked in bug 1356138. I don't know if this is "invalid" or "wontfix", but either way, chances are we do not want to put back a sync IPC in order to throttle content process and give us better talos results on session restore. I'll start with wontfix on both these bugs, make a change if it should be something else.