Re: [avr-libc-dev] Re: [bug #13241] ADHSM missing from iom16.h

[E. Weddington]

Joerg Wunsch wrote:

> As E. Weddington wrote:
>
>  
>
> > > ATmega16 datashet (2466K?AVR?04/05) on page 342:
> > >
> > > Changes from Rev.
> > > 2466E-10/02 to Rev.
> > > 2466F-02/03
> > >
> > > ...
> > > 8. Removed ADHSM completely.
> > > ...
> > >      
>
>  
>
> > Thanks, Anatoly!
> >    
>
> Btw., I've asked address@hidden for the reasoning behind that.
> They replied to me, but so far did not respond to my request
> for permission to quote their reply verbatim here.
>
> So in my own words, that bit was once meant to feed additional power
> into the analog comparator so it would get faster, but they eventually
> noticed the comparator was even fast enough without that measure.  In
> effect, the ADHSM bit turned into a ``burn more power'' bit only, so
> they withdrew it.  According to their reply, the presence of that bit
> in the current AT90CAN128 docs is a mistake, and scheduled for removal
> as well.
>
> In that light, I think we should change our header files to remove the
> bit as well, 
Do you mean just the can128 header? Or are there others that you're 
thinking of?


> so any attempt to still use it will point the people onto
> it by a due compiler error.  The removal should IMHO be done by
> commenting it out and stating the reason, so everybody gets a clear
> idea that he really doesn't want that bit anymore at all.
>
>  
Would you mind filling out a bug report so we don't forget? ;-)

Eric