Re: [PATCH] user32: Add a timeout when sending WM_WINE_DESTROYWINDOW

[Alexandre Julliard]

Andrew Eikum  writes:
> Without this change, the test will deadlock in DestroyWindow waiting
> for the other thread to process the WM_WINE_DESTROYWINDOW message.
>
> Note that this adds about 1 second of runtime to the test, waiting for
> the SendMessage to time out. Windows does not time out like this.
> Another option would be to use PostMessage, but that seems like a
> larger change.

Your test doesn't deadlock here, and I'm not sure I see why it should,
since the thread is processing messages. Do you have a backtrace of the
deadlock?