[avrdude-dev] ATmega169 efuse wrong?

[Joerg Wunsch]

Holm Tiffe pointed out to me that he was unable to program his desired
BODLEVEL into the extended fuse byte of an ATmega169.  I think the
datasheet is self-contradictory here: in the table describing the
efuse bits, it effectively says bit 1..3 are to be programmed, while
in the table summarizing the serial downloading commands, only bit 0 and
1 are marked as input.

I tried the attached patch (by now only on an STK500/STK502, not
directly via a parallel port attached ISP dongle), and it seems I can
successfully program the efuse bits then to whatever BODLEVEL I see
fit.

Can anyone deny/confirm this?

I've also filed a support request to Atmel for this.

-- 
J"org Wunsch                                           Unix support engineer
address@hidden        http://www.interface-systems.de/~j/