You can get away with an awful lot of bogus starring of Android failures, right up until you get caught, merge around bustage, and wind up closing m-i, m-c and fx-team over it. (I already deleted the bogus starring, but rest assured, we're all guilty.) https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&filter-searchStr=8ee582708818b64631bc56ffa04b7b1138e1e503 https://treeherder.mozilla.org/#/jobs?repo=mozilla-central&filter-searchStr=1ca1e6c58c395c410a157bf19bbc12cbaec56527 https://treeherder.mozilla.org/#/jobs?repo=fx-team&filter-searchStr=a665adcc8cbca924f012ec7bd55dd3f2653b80fe I'm currently leaning toward fx-team as the start of it, though retriggering to find out won't be much fun since we're currently massively backlogged.
There's some discussion in bug 1161150 about crashes in testInputUrlBar.
Silly of me not to look in an obvious misstar bug for it, thanks! ;)
https://treeherder.mozilla.org/#/jobs?repo=fx-team&filter-searchStr=a665adcc8cbca924f012ec7bd55dd3f2653b80fe&fromchange=eac6ac60b5e6&tochange=e05936099653 I honestly don't know what to do about that. It went from a completely unacceptable 20% crash to a completely unacceptable 90% crash on what can only be https://hg.mozilla.org/integration/fx-team/rev/e05936099653 since everything else in that range is browser/.
In the end, apparently fixed by backing out bug 1137483.