[En-Nut-Discussion] atmega2560 support

[Dusan Ferbas]

Hi Thiago,

we put a lot of effort to run code on ATmega2561.
Currently now, nearly all features works, but I am convinced it will 
not work once we reach 128kB flash boundary.

I do not recommend to use WinAVR 20080610. With this, our code does 
not run. Worse, code size is greater, even they claim, compiler optimizes more.
We use gcc version 4.2.2 (WinAVR 20071221).
You can see my discussion at 
http://lists.gnu.org/archive/html/avr-gcc-list/ (look for eicall, eind, 2561).

I discussed bootloader issue with Andy Hutchinson and realized, that 
compiler does not support wider (>16 bits) function pointers.
Compiler generates eicall instructions (EIND + Z is used for 3 byte 
calls), but it never handles EIND register.
So if you have indexed calls in your application, it will work only 
in lower 64kW.
(Unless you manually do call with asm)

BTW, anyone experienced with EIND & eicall ?
(because it works for us in bootloader area when entered after power 
up, but not if we call it from application)

>At 21:41 6.6.2008, hutchinsonandy at aim.com wrote:
>CALL  can reach all 4M of address space.
>
>EICALL is special version of ICALL - which would be appropropriate 
>for non-constant function pointers. And this would be the area in 
>which near/far pointer support would be useful. However, the true 
>functions pointers inside compiler are only 16bits, so its not a 
>trivial change.
>
>But your code has constant pointer. So CALL  should be ok.
>
>Andy

Dusan

------------
At 17:01 10.7.2008, "Thiago A. Correa" <thiago.correa at gmail.com> wrote:
>Hi,
>
>    I've done some of the work required to support Atmega2560 (mostly
>preprocessor defines thanks to 2561 support already in the tree). But
>I'm not able to "boot" the board to the user code. If I try to run the
>code thru JTAGICE mkII it just sits there blinking without ever
>reaching any code as it does with atmega128's.
>...

Dusan