unionfs related patch

[David Schultz]

On Thu, Sep 04, 2003, Marc G. Fournier wrote:
> 
> 
> On Thu, 4 Sep 2003, David Schultz wrote:
> 
> > On Mon, Sep 01, 2003, Alexandr Kovalenko wrote:
> > > Hello, Alexandr Kovalenko!
> > >
> > > On Mon, Jun 23, 2003 at 11:19:27AM +0300, you wrote:
> > >
> > > > >   David Schultz, in his spare time, has been working through some of the
> > > > > issues I've been able to 'tweak' in the unionfs code ... as he is
> > > > > currently working on forward-patching it to -CURRENT right now, he can't
> > > > > commit the code to the -STABLE tree ... in order to allow others using
> > > > > unionfs to test the patch (I've been running it a few weeks now on a very
> > > > > heavily loaded system), with his permission, I've posted the patch for
> > > > > download at:
> > > > >
> > > > > 	http://www.hub.org/~scrappy/unionfs
> > > >
> > > > Looks like it pretty stable. I'm running ~10 days with this patch - no
> > > > even single panic. Thank you!
> > >
> > > I was too optimistic. It panics on high loads after 3-10 days.
> > > Is there any progress in work on this patch?
> >
> > The only thing I know to be wrong with the patch is that it looks
> > like I forgot to MFC it.  Oops.  Unfortunately, there's plenty
> > wrong with unionfs, even with the patch.
> 
> Huh?  You MFC'd it ... -STABLE from Saturday has the memory allocation
> stuff still in it:
> 
>   UNION mount    29     1K      1K204800K       30    0     0  32
>        undcac     0     0K      1K204800K 28159584    0     0  16
>        unpath  6086   104K   1853K204800K 26015006    0     0  16,32,64,128

Heh.  Never mind, I'm not that forgetful after all.  When I saw
your link to the patch for -STABLE, I thought I must have missed it.
But it looks like that was merely a quote of an old email.