[avrdude-dev] [bug #25755] avrdude > 5.2 uses wrong avr910_devcode

[Joerg Wunsch]

Update of bug #25755 (project avrdude):

                  Status:                    None => Invalid                
             Assigned to:                    None => joerg_wunsch           
             Open/Closed:                    Open => Closed                 

    _______________________________________________________

Follow-up Comment #1:

avr109 aka. "butterfly" is supposed to ignore the historic
AVR910 device code given.  The validity of the target part
is supposed to be handled by matching the signature bytes,
and that's exactly what AVRDUDE does.

For historical reasons, AVRDUDE selects the first device
code presented by the bootloader, just to send something
at all.  After all, this is a bootloader protocol, and
bootloaders are hardly expected to be able to handle more
than a single device code anyway.

If, for some reason, that doesn't really fit your particular
bootloader, you are still free to resort to the classic
AVR910 protocol instead.  As the device code mess could never
resolved in a way that would satisfy all interested parties,
the option -x devcode=XXX has recently been added there, as
a method to override the device code selection taken from
avrdude.conf straight from the commandline.

Anyway, you are *strongly* suggested to send all the devcodes
to /dev/null, and rely on the device IDs (signatures) instead.
AVR910 single-byte device codes (and STK500 v1 device codes,
for that matter) have been a design failure from day #1.

    _______________________________________________________

Reply to this item at:

  <http://savannah.nongnu.org/bugs/?25755>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.nongnu.org/