we need to be able to test the performance of a broad range of specific actions in the browser, such as opening a new tab, or opening the history menu. there are two parts to resolving this: 1) the test framework some exploratory work was done in bug 451607 using mochitests and reporting via test comments. it's not a realistic test of stability of the numbers though, since they're being run in the midst of the actual unit tests. we should instead have a separate box that runs *only* the mochitests for perf. basically it's a new type of mochitest, maybe we could prefix with perf_ instead of browser_ or test_, or something like that? 2) reporting seems like there's two ways to go about this: - have a separate instance of perfomatic, since all these little test scores will muddy the test list, generally making things slower and more cumbersome to dig through - roll all the individual test scores up into a single aggregate result i'd prefer a separate performatic instance, so that we can track each test individually.