[Icc-mot] GNU to ICC

[Oskar Atkinson]

That would give me quite a bit more available memory.

As the original code, compiled with GNU, did fit in a MC9S12C64 - I hope
and expect that the same code ( once finally compiling) compiled with
ICC will fit as well.

I still am puzzled by the ton of errormessages:

> !E ms2_extra_main.o(2749): Code address 0:0x400b already contains a
> value
> !E ms2_extra_main.o(2749): Code address 0:0x400c already contains a
> value
> Counting up to
> !E ms2_extra_main.o(2749): Code address 0:0x6186 already contains a

Basically for every byte I am getting a line. 
IMHO, it has to have some cause - which I am not aware of.


( Using my standard dev environment Visula Objects - I can create such a
situation of 1000nds of errors. But there I know how to solve it .. )


Oskar

  

-----Original Message-----
From: icc-mot-bounces at imagecraft.com
[mailto:icc-mot-bounces at imagecraft.com] On Behalf Of Edward Karpicz
Sent: Wednesday, February 20, 2008 1:14 PM
To: Discussion List for ICC08/11/12/16 users. You do NOT need
tosubscribetoicc-announce if you are a member of this.
Subject: Re: [Icc-mot] GNU to ICC

I forgot that C family members, different compared to D's and A's, can
have linear nonpaged memory area larger than 48k. Good news for you,
isn't it? I didn't try C64 myself, but if I understand datasheet
properly, then you can have up to 64-4=60k of linear nonpaged flash. If
you didn't read in the docs about internal memories priorities, then
please note that RAM takes precedence over flash. It means when RAM and
flash addresses do overlap, then RAM is visible and piece of flash,
covered by RAM block, isn't visible. 
And the next rule - register block has the highest priority. That's why
64-4k. INITRM register out of reset is set to 9, so RAM block is at
0x0.0xFFF. INITRG register defaults at 0, so register block occupies
0x0.0x3FF addresses. So out of reset you have 2k of registers at
0x0.0x3FF, 3k of RAM at 0x400.0xFFF, and 60k of flash at 0x1000.0xFFFF.
If you need all 4k of RAM, then you will have to remap RAM for example
to 0x1000.0x1FFF, loosing extra kB of flash.

So maybe try changing program memory setting to 0x1000.0xFFFF. If it
won't fit 60k, then I guess you need more flash and paged memory.


Edward

P.S. you may have problems flashing compiled 0x1000.0xFFFF image due to 
possibly lack of support from debugger/programmes. But this is solvable 
converting S records to proper format.


Oskar Atkinson wrote:

> Well, looking different now ... :-)
>
> However I need to bug you again...
>
>>>
> How much space do you need for your constants? Could your constants
and
> all the code into 48k? It would be the best for you, just use nonpaged
> 0x4000.0xFFFF area.
> <<
>
> Seems it not fitting in 48k,
>
> icc12w -o MSIIETXTRA-ICC-2008 -btext:0x4000.0xFFFF -bdata:0x0400
> -dinit_sp:0x1000 -fmots19 @MSIIETXTRA-ICC-2008.lk
> !E <library>(200): area 'text' not large enough
> want size 24967
> lo 49487 hi 65535 size 16049
>
> And then I am getting a gazillion errors:
>
> !E ms2_extra_main.o(2749): Code address 0:0x400b already contains a
> value
> !E ms2_extra_main.o(2749): Code address 0:0x400c already contains a
> value
> Counting up to
> !E ms2_extra_main.o(2749): Code address 0:0x6186 already contains a
> value
>
>
> Oskar
>
> _______________________________________________
> Icc-mot mailing list
> Icc-mot at imagecraft.com
> http://dragonsgate.net/mailman/listinfo/icc-mot
> 

_______________________________________________
Icc-mot mailing list
Icc-mot at imagecraft.com
http://dragonsgate.net/mailman/listinfo/icc-mot