[Webkit-unassigned] [Bug 114696] Wrong use of REALTIME clock causes pages to fail loading

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=114696





--- Comment #10 from Joe Mason <jmason at blackberry.com>  2013-04-18 12:02:41 PST ---
(In reply to comment #8)
> Wait, but why is this correct?

As far as I can see, m_nextFireTime is only used to pass to waitForMessageFilteredWithTimeout in runInMode:

    if (waitMode == WaitForMessage)
        absoluteTime = (predicate.isDefaultMode() && m_sharedTimer->isActive()) ? m_sharedTimer->fireTime() : MessageQueue<Task>::infiniteTime();
    MessageQueueWaitResult result;
    OwnPtr<WorkerRunLoop::Task> task = m_messageQueue.waitForMessageFilteredWithTimeout(result, predicate, absoluteTime);

>From there it's passed to PlatformCondition::timedWait... which compares it against currentTime() in the pthreads implementation. So this is not correct.

It does seem like an error in PlatformCondition, though - it seems to me that it should not be affected by changes to the clock. But changing that would involve changing a lot of calling code.

Glad you caught that! I thought I'd verified that each changed in the patch was only used in a direct comparison.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.