Hi,

So I'm back from travel, I'll be looking into those thing shortly.

As for notifying Atmel of mistake/errors/inconsistencies in XML files
I thing it will be harder than what you make it sound like. Using the
librairy, we load the XML in ram as an object tree. When I walk the
tree, I can find errors but I cannot relate back to the XML line.

Also, I'm not sure what should be flaged as an error. It seems to me
that there is 2-3 different pattern that they follow for pin
description. Are they all right? In AVR Studio XML parser, they for
sure run into the same problem that I do.

I'll update the header soon this week.

Frederic

2009/4/1 Weddington, Eric <[email protected]>:
>
>
>> -----Original Message-----
>> From:
>> [email protected]
>> [mailto:avr-libc-dev-bounces+eric.weddington=atmel....@nongnu.
>> org] On Behalf Of Frédéric Nadeau
>> Sent: Tuesday, March 31, 2009 3:59 PM
>> To: [email protected]
>> Subject: Re: [avr-libc-dev] Re: [bug #25300] Additional i/o port names
>>
>> 2009/3/31 Weddington, Eric <[email protected]>:
>> > Ideally, we need both: we need to fix the patch to the
>> header files so that the definitions are in uppercase, but we
>> need to generate a list of what was wrong with the XML files
>> to give to Atmel.
>> >
>>
>> Making the script to generate all in upper case is a piece of cake.
>> However having the script to report all the error is more or less
>> complicated. I'll see what I can do about that
>
> Something else I just found out:
> I was testing my generator script with the ATmega6490 and it reported back 
> that there was duplicate definitions with the new pin definitions:
>
> #define SEG33_DDR   DDRJ
> #define SEG33_PORT  PORTJ
> #define SEG33_PIN   PINJ
> #define SEG33_BIT   1
>
> #define SEG33_DDR   DDRG
> #define SEG33_PORT  PORTG
> #define SEG33_PIN   PING
> #define SEG33_BIT   3
>
> I checked and it seems that the XML file is incorrect. :-P
>
> This duplicated data is in your patch too, unfortunately.
>
> So when you generate the new definitions, you'll need to check to make sure 
> that there are no duplicate definitions. If there are, you'll have to verify 
> which one is correct and probably fix the header by hand.
>
> Eric Weddington
>


_______________________________________________
AVR-libc-dev mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/avr-libc-dev

Reply via email to