Re: bdrv_drained_begin deadlock with io-threads

[Dietmar Maurer]

> > But, IMHO the commit is not the reason for (my) bug - It just makes
> > it easier to trigger... I can see (my) bug sometimes with 4.1.1, although
> > I have no easy way to reproduce it reliable.
> > 
> > Also, Stefan sent some patches to the list to fix some of the problems.
> > 
> > https://lists.gnu.org/archive/html/qemu-devel/2020-04/msg00022.html
> > 
> > Does that fix your problem?
> 
> It seems to fix it, yes. Now I don't get any hangs any more. (Also, I
> guess this means that this day was essentially wasted because I worked
> on a problem that already has a fix... *sigh*)

It is not wasted, because we now know that you can't reproduce the bug we 
observe here. 

We are hunting those AIO bugs for about three weeks now, and it looks we still 
have
not found all of them ...

Also, it would we great if somebody reviews the patches from Stefan. It would 
be great to
get them in before 5.0.0.