[Webkit-unassigned] [Bug 167003] REGRESSION(r202615?): [GStreamer] ASSERTION FAILED: isMainThread() in WebCore::BuiltinResourceHandleConstructorMap& WebCore::builtinResourceHandleConstructorMap()

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=167003

Michael Catanzaro <mcatanzaro at igalia.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[GStreamer] ASSERTION       |REGRESSION(r202615?):
                   |FAILED: isMainThread() in   |[GStreamer] ASSERTION
                   |WebCore::BuiltinResourceHan |FAILED: isMainThread() in
                   |dleConstructorMap&          |WebCore::BuiltinResourceHan
                   |WebCore::builtinResourceHan |dleConstructorMap&
                   |dleConstructorMap()         |WebCore::builtinResourceHan
                   |                            |dleConstructorMap()

--- Comment #3 from Michael Catanzaro <mcatanzaro at igalia.com> ---
Carlos, I think you broke this when trying to fix HLS in r202615. In ResourceHandleStreamingClient::ResourceHandleStreamingClient, you create the ResourceHandleStreamingClient thread, in which you immediately call ResourceHandle::create. But this causes an immediate crash in debug builds because ResourceHandles must only be created on the main thread.

I am really confused why we did not notice this until now, because creating a ResourceHandle off the main thread has been an immediate assert for a long time. It's not some new restriction.

-- 
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/20170129/5e43f25d/attachment.html>