Telemetry: https://mzl.la/2rp2aeP

Andreas, I think the telemetry collection code in bug 1322184 may be the reason for this regression... The good news is that if this is the case, AutoCollectVsyncTelemetry automatically turns itself off for the release channel, so we can just ignore the regression. But we should verify whether this is really the regression we are looking at. I would like to disable this probe for a few days and watch INPUT_EVENT_RESPONSE_MS to see if it drops down and in that case, mark this bug as WONTFIX and move on!

Sorry, shouldn't have set the checkin-needed flag there. Don't know what happened.

What I should've said is that we should probably do this as well.

Comment on attachment 8869396 [details] [diff] [review] 0001-Bug-1366156-Temporarily-stop-recording-vsync-message.patch Yeah makes sense. I'll fold this into my patch to make it easier to back the whole thing out after a few days. Thanks!

Pushed by eakhgari@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/dfdebc342d2e Temporarily disable the collection of content js delay event telemetry probes to investigate whether they are the cause of child process INPUT_EVENT_RESPONSE_MS regressions; r=farre

Comment on attachment 8869307 [details] [diff] [review] Temporarily disable the collection of content js delay event telemetry probes to investigate whether they are the cause of child process INPUT_EVENT_RESPONSE_MS regressions It doesn't look like this has caused any measurable change in the INPUT_EVENT_RESPONSE_MS telemetry probe. I'm going to revert this patch now...

(In reply to :Ehsan Akhgari (super long backlog, slow to respond) from comment #8) > Comment on attachment 8869307 [details] [diff] [review] > Temporarily disable the collection of content js delay event telemetry > probes to investigate whether they are the cause of child process > INPUT_EVENT_RESPONSE_MS regressions > > It doesn't look like this has caused any measurable change in the > INPUT_EVENT_RESPONSE_MS telemetry probe. I'm going to revert this patch > now... https://hg.mozilla.org/integration/mozilla-inbound/rev/1658b1c63226

Merge of backout: https://hg.mozilla.org/mozilla-central/rev/1658b1c63226

Backed out bug 1352889 today to see whether it is the cause.

The backout of bug 1352889 didn't make any difference. I'm going to reland it, but that means you need to look for a different culprit here.

See bug 1362094 comment 25. We decided this bug is INCOMPLETE due to the limitation of the probes.

Removing leave-open keyword from resolved bugs, per :sylvestre.