In bug 989533 we started running "mochitest-e10s" on some trees. That runs all of our Mochitests (modulo a bunch that have been disabled until we can fix them to pass). We have a handful of "test_ipc.html" sub-harnesses in the tree that explicitly run tests in a content process (I believe IndexedDB was the first mover here). I believe that mochitest-e10s coverage is equivalent to the test_ipc.html coverage. Once we get m-e10s green enough to be visible on all trees, and have all the tests that are currently running in test_ipc subharnesses running, we ought to be able to remove the test_ipc bits in favor of the coverage from m-e10s. Anyone disagree?

Totally agree. WebRTC uses test_ipc.html at dom/media/tests/ipc. And I would love to get rid of them.

Who would need to sign off on this?

I think we'd just need someone working on e10s to verify that we get equivalent test coverage here.

Well, I assume we'll need e10s tests running in opt and debug on all platforms. We don't have that yet.

(In reply to Bill McCloskey (:billm) from comment #4) > Well, I assume we'll need e10s tests running in opt and debug on all > platforms. We don't have that yet. Do we have a ticket for tracking the progress towards that goal?

will solving bug 1194533 allow us to disable test_ipc tests?