Re: [avr-libc-dev] __vector_N off-by-one?

[Joerg Wunsch]

As Bernd Trog wrote:

> > The base for the counter differs by 1 between both: avr-libc counts
> > starting at 1 for the first interrupt vector, the AVR datasheets start
> > counting at 1 for the reset vector.

> Would it break existing code if you make the avr-libc __vector
> consistent with the AVR data sheets?

Dunno.  At least, it would require to consistently change it in both,
avr-libc and avr-binutils, which will end up in a maintenance nightmare
since both packages are developed independently.

I see no reason though; symbols starting with two underscores are
meant to be for internal (library) use only, and do not consitute a
public interface anyway.  It would be more important to find a way
to spell-check the vector names IMO, since this is a frequent error
users stumble across.
-- 
J"org Wunsch                                           Unix support engineer
address@hidden        http://www.interface-systems.de/~j/