Now that we are getting closer to be able to turn on the e10s pref we can start running tests and trying to get them green. I'll hang various fixes off of this. The starting point here is with patches from bug 1459209 and bug 1462069.

There will be a lot of orange in here. It at least still has bug 1462464 in it: https://treeherder.mozilla.org/#/jobs?repo=try&revision=1a83213558ea6b261216f1b4a4c2c8df5c6e8a1e

My current work-in-progress patch queue has green WPT and mostly green mochitests. Main failing mochitests are things like devtools, browser chrome tests, etc. https://treeherder.mozilla.org/#/jobs?repo=try&revision=5e8e6d046a34f85cd55039618edfa3a3e6325cc1

Valentin, as I discuss in comment 0 I want to preserve channel creation and async open time across internal redirects. Our current practice of resetting the start of the channel time on every internal redirect produces overly optimistic numbers since it ignores real network stack execution time. Also, I need this in order to properly track time across the additional internal redirects used by the new service worker parent intercept. There is a test that verifies that the service worker intercept time is accurately reported. Currently its missed in this mode without a change like this. https://treeherder.mozilla.org/#/jobs?repo=try&revision=2712e428fd77d9772075fd7bbd2e0ea5d5ea4f7f

Thanks for the review. I accidently put this patch and review on the wrong bug. I'll move it to the correct bug before landing.

Comment on attachment 8986904 [details] [diff] [review] Propagate channel creation and async open time across internal redirects. r=valentin Moved to bug 1470286.