On Mon, 29 Jan 2007 07:50:41 -0500
"Blair Sadewitz" <blair.sadewitz@gmail.com> wrote:

> BTW, I thought I'd follow up with some more important information on
> BUFQ_PRIOCSCAN.  Yamt@ informaed me that the reason it is still deemed
> "experimental" is that it interacts poorly with the UVM system such
> that--with certain kinds of interactive activities--performance may
> suffer.

um. thats exactly the same as the baseline default, and the previous
alternate.

ie. *all* of the UVM system interactions can react badly to interactive
performace!

> I do not recall what exactly the problem was, as I can't find the
> email right now.
> 
> It is worthy of mention that when I disabled BUFQ_PRIOCSCAN, I no
> longer had audio skipping problems.  Thus, unfortunately,
> BUFQ_PRIOCSCAN isn't a panacea. :(

No. Right now, for me, modulo not needing audio, its heaps better than
the alternatives, and I *am* on a box which depends on interactive
performance, and does king-hit disk I/O from time to time.

YMMV

-G

> 
> Regards,
> 
> --Blair
> 
> -- 
> Support WFMU-FM: free-form radio for the masses!
> 
> <http://www.wfmu.org/>
> 91.1 FM Jersey City, NJ
> 90.1 FM Mt. Hope, NY
> 
> "The Reggae Schoolroom":
> <http://www.wfmu.org/playlists/RS/>
>