avrdude-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[avrdude-dev] [bug #49898] Config file error for many parts


From: Owen Duffy
Subject: [avrdude-dev] [bug #49898] Config file error for many parts
Date: Wed, 21 Dec 2016 13:04:46 +0000 (UTC)
User-agent: Mozilla/5.0 (Windows NT 10.0; rv:50.0) Gecko/20100101 Firefox/50.0

Follow-up Comment #2, bug #49898 (project avrdude):

I read most of the long discussion.

For as long as I can remember, the Atmel tools have required the user to
specify ONLY the implemented bits for the example case I gave, three bits.
They do not state in the datasheet that the user must specify unimplemented
bits at 1, but they do hint that they will read back as one (because being
unimplemented, the should always be 'unprogrammed'.

In Atmel's tools if I select BOD=2.7 from the feature selection, it calculates
and shows efuse=0x05, and if programmed, that WILL verify.

That is all rational.

It seems that recently, AVRDUDE's direction is to mask out ONLY three bits
from user supplied efuse parm (for this part), but include 8 bits read back
from the part for verification. Why verify bits that the user did not
specify.

AVRDUDE has changed to strike out on its own, it requires a different efuse
specification than the Atmel tools display... not for programming the fuse
correctly, but purely for verification.

Most tools follow Atmel's lead, eg http://www.engbedded.com/fusecalc/ .

I think you have made a mistake, the solution for users who want consistency
with Atmel and compatibility with all the existing scripts, makefiles and
project documentation seems to be v6.1.


    _______________________________________________________

Reply to this item at:

  <http://savannah.nongnu.org/bugs/?49898>

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




reply via email to

[Prev in Thread] Current Thread [Next in Thread]