[Webkit-unassigned] [Bug 144824] New: [GTK] WTF_WorkQueue.Simple test fails due to assumptions about reference count

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=144824

            Bug ID: 144824
           Summary: [GTK] WTF_WorkQueue.Simple test fails due to
                    assumptions about reference count
    Classification: Unclassified
           Product: WebKit
           Version: 528+ (Nightly build)
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Tools / Tests
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: mcatanzaro at igalia.com

**FAIL** WTF_WorkQueue.Simple

../../Tools/TestWebKitAPI/Tests/WTF/WorkQueue.cpp:57

Value of: initialRefCount

  Actual: 2

Expected: 1

This is not a bug: we need the second ref to keep the WorkQueue alive in the lambda function where the main loop runs in another thread, in WorkQueue::platformInitialize:

RefPtr<WorkQueue> protector(this);
m_workQueueThread = createThread(threadName, [protector] {
    g_main_context_push_thread_default(protector->m_eventContext.get());
    g_main_loop_run(protector->m_eventLoop.get());
    g_main_context_pop_thread_default(protector->m_eventContext.get());
});

Since we need to keep the WorkQueue alive in another thread, I think the only option is to change the test, not the implementation. CC Brent since he wrote this test. We could guard the reference count checks with #if !PLATFORM(GTK), or just remove the checks completely.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20150509/faf6e01e/attachment.html>