Re: [avrdude-dev] Bad programming speed on AT90S8535 using AVRdude

[Theodore A. Roth]

On Tue, 18 Nov 2003, Brian Dean wrote:

> On Tue, Nov 18, 2003 at 05:45:02PM -0800, Theodore A. Roth wrote:
> 
> > I've digging on google for some reason why and found nothing. I also
> > tried tweaking the ppdev stuff in avrdude and came up with nadda.
> 
> Well, I updated my P-III 300 to the latest FreeBSD code, exactly what
> is running on my faster Athlon and it still programs in 39 seconds -
> no change from before.
> 
> I did notice that during the programming, according to 'top', avrdude
> went to 93% CPU utilization on the slower system, and marched up to
> 80% utilization on the faster system and looked like it would keep on
> rising but the job finished at that point.  So maybe the parallel port
> bit banging is more CPU intensive than I first thought, and have a
> faster system makes a big difference.

Does FreeBSD allow for multiple processes to share the ppi interface at the
same time? The ppdev driver does which is why linux needs the ppi_claim
macro. I'm wondering if that may account for the extra overhead which slows
linux down... I tried to do a PPEXCL before the PPCLAIM to test this
hypothesis, but couldn't get it to work and the docs are a bit sparse there.

Ted Roth