On Mon, Jun 4, 2012 at 2:53 AM, Alexandre Julliard wrote:
> ...
> I don't think that you can have an async queue that outlives its file
> descriptor.
Well, maybe I screwed something up - but when I tried to access the
descriptor like that it caused segfaults in some of the tests. What I
found wa
"Erich E. Hoover" writes:
> Well, maybe I screwed something up - but when I tried to access the
> descriptor like that it caused segfaults in some of the tests. What I
> found was that neither create_async_queue or create_async increment
> the refcount on the file descriptor, but create_async do
"Erich E. Hoover" writes:
> Sorry I didn't reply to this sooner, I've been at a wedding at just
> got back to the land of the Internet. I believe that it really does
> need to be stored in the queue, otherwise in the case that the async
> operation has not completed when the file descriptor is c
On Fri, Jun 1, 2012 at 8:57 AM, Alexandre Julliard wrote:
>
> "Erich E. Hoover" writes:
>
> > Real Name:
> > Erich Hoover
> >
> > Description:
> > It seems that the completion information should be associated with
> > the async queue, rather than the async object, since the completion
> >
"Erich E. Hoover" writes:
> Real Name:
> Erich Hoover
>
> Description:
> It seems that the completion information should be associated with
> the async queue, rather than the async object, since the completion
> information for a file handle can be updated after an async IO has
> been que
