https://tbpl.mozilla.org/php/getParsedLog.php?id=19744999&tree=Mozilla-Inbound Rev3 Fedora 12 mozilla-inbound pgo test mochitest-3 on 2013-02-14 11:33:22 PST for push 32327a07b56a slave: talos-r3-fed-064 424 INFO TEST-START | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html 425 INFO TEST-INFO | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html | Step 0: Instantiate a Peer Connection 426 INFO TEST-INFO | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html | Step 1: Get User Media for Audio and Video 427 INFO TEST-INFO | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html | Step 2a: Add stream 428 INFO TEST-INFO | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html | Step 2b: Create Session Description 429 INFO TEST-INFO | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html | Step 2c: Set Remote Description 430 ERROR TEST-UNEXPECTED-FAIL | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html | Test timed out. args: ['/home/cltbld/talos-slave/test/build/bin/screentopng'] SCREENSHOT: <see log> 431 INFO TEST-END | /tests/dom/media/tests/mochitest/test_peerConnection_bug840344.html | finished in 325105ms

Hmm...looks like we are failing when we call setRemoteDescription with a session description object with that custom SDP.

We also appear to only be failing on fedora opt builds.

(In reply to Jason Smith [:jsmith] from comment #3) > We also appear to only be failing on fedora opt builds. and pgo builds.

Not actionable without more data from those logging bugs

Reliably(?) stopping at 2c SetRemoteDescription

Still always times out after 2c, SetRemoteDescription() - success and error callbacks never fire.

Any TBPL stars after this comment should contain useful logging information that isolates this problem to a smaller part of the system.

So we've had a bunch of reports since we added the debugs. Do we know anything more about this? Is this something someone else can help on? (jib)?

FWIW, I've added additional annotation explicitly aimed at diagnosing intermittent timeout bugs as part of trying to track down bug 853858. That patch landed on m-i just a few minutes ago. If it stays in the tree (and I expect it will), then future runs of this will have far better breadcrumbs for figuring out where our call went. See Bug 853858 comment 3 for details. I believe Maire is going to ping jib to see if he has some cycles to dig into this bug.