Re: [avrdude-dev] [patch #4925] Check part's signature against expected value

[Joerg Wunsch]

As Brian Dean wrote:

> Perhaps we can also lift the restriction that the part number be
> supplied on the command line since it can now match up the signature
> ID with the config file entry.

I've also thought about that, it would be really nice, yes.

>  Would that be a circular dependency, though?  Some programmers need
> to know what the device type is before entering program mode, and
> you can't get the device type, until you enter program mode and
> query the chip.

AFAICT this only applies to the AT90S1200 which is now discontinued
anyway.  So the only dependency is that you cannot automatically
detect the presence of an AT90S1200 so that one *must* be specified on
the command-line.  Likewise, for any AVR with a broken chip ID (which
could happen from time to time, though I haven't observed that anymore
lately), the MCU type must be supplied in advance, and obviously, the
-F flags needs to be given to override AVRDUDE's chip ID check.

-- 
cheers, J"org               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/                        NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)