Summary: | REGRESSION(r176687): Caused webkit2 test assertion failure crashes in WebKit::WebPage::willStartDrag() | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Mark Lam <mark.lam> | ||||
Component: | WebKit2 | Assignee: | Nobody <webkit-unassigned> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | ap, mitz | ||||
Priority: | P1 | Keywords: | LayoutTestFailure, Regression | ||||
Version: | 528+ (Nightly build) | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Mark Lam
2014-12-03 09:22:43 PST
Updated TestExpectations to expect crashes for these tests in r176723: <http://trac.webkit.org/r176723>. I think that a Skip expectation would be more appropriate. Seven tests that crash every time make bots substantially slower, and I suspect that the high system load associated with creating crash logs may destabilize other tests too. So, the better expectations would be (note how they can be limited to debug builds only): webkit.org/b/139224 [ Debug ] fast/events/clear-drag-state.html [ Skip ] ... Also, crashes make running tests locally very annoying. Updated TestExpectations to skip tests per Alexey's instructions in r176726: <http://trac.webkit.org/r176726>. Created attachment 242503 [details]
Wait for the UI process to begin dragging if needed before handling a synthetic synchronous mouse event
Fixed in <http://trac.webkit.org/r176728>. |