Subject: Re: possible arm32 fp tuning project.
To: Erik E. Fair <fair@clock.org>
From: Scott Reynolds <scottr@og.org>
List: port-arm32
Date: 10/11/1998 21:47:52
On Sun, 11 Oct 1998, Erik E. Fair wrote:
> At 7:33 -0700 10/11/98, Herb Peyerl wrote:
>
> >mmmmm, I'd love to have some sort of hack that would allow my shark to
> >decode mp3 in realtime... I don't mind running something that's not
> >in-tree...
>
> Me, too, but my impression from reading this list was that the problem was
> the arm32 port's interrupt structure, and resulting interrupt response
> latency [...]
Herb's correct; it's the floating point performance. We proved this with
the June snapshot by replacing libc with the faster-fp version available
on ftp.netbsd.org.
--scott