On 12.03.13 12:53, Thomas, George wrote:
>
> const int __memx *ramVar = &k;
...
> When I checked the linker scripts, all __flashN's are going into the same
> location.
Oh-oh, I'd up till now only read the text, not the code snippet.
We handle explicit __flashN's in the new linker script, but lack of
response has held up implementation of __memx.
Could you please show what the compiler does with __memx?¹
If it places successive hunks of __memx into successive .flashN input
sections, then it is already handled by the new script, I anticipate.
So far, we are lacking any concrete requirements for __memx, AFAICT.
If you read the two long related threads, you'll see that we haven't
gotten around to clarifying how that one big address space should
behave.
Erik
¹ It's nearly midnight here - too late to start experimenting with that.
--
Don't worry about avoiding temptation -- as you grow older, it starts
avoiding you.
- The Old Farmer's Almanac
_______________________________________________
AVR-GCC-list mailing list
AVR-GCC-list@nongnu.org
https://lists.nongnu.org/mailman/listinfo/avr-gcc-list
