WINAVR : next release..?

Go To Last Post
8 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am just curious: Does anybody know, if there is an update in the pipeline?

There were some rumors about a new WINAVR with GCC-4.1.1 and support for 256k Devices. But now, there seems to be no life sign from Eric since September..?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well the problem is that there is no support in AVR Studio for 32bit dwarf objects.

I compiled my own AVR-GCC 4.1.1, works fine, but cannot use it for debugging/simulating in AVRstudio, since the new debug-format is not supported by AVRstudio.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well! Then, it would be from big interest, if Atmel does have any plans to support 32bit dwarf with AVR-Studio in the near future?

Does anbody know more details, or shall I send a request to Atmel?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I can't remember if it was Torleif or Roland but one of the Studio developers posted here to say they were working on the additional ELF parsing in Studio to match the support in GCC so presumably this is all just a question of co-ordination between the various Atmel employees (now that Eric is on the "inside")

Cliff

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hmmm..

Quote:
pfleury wrote:
Well the problem is that there is no support in AVR Studio for 32bit dwarf objects.

What would be the symptoms of this lack of support if, say, an individual builds an .elf (+dwarf2) with 4.1.1 and tries to load it into Studio?

Because, I've built avr-gcc 4.1.1 with the dwarf patch, stock binutils 2.17 and avr-libc 1.4.5, and I seem to be able to build with them and can debug from within AVR Studio 4.12 build 498.

I'm getting some stack underflow messages, but otherwise (and regardless of those) my program runs correctly in the simulator.

I don't have any profound understanding about the tools. Basically I just followed these instructions.
http://homepage.hispeed.ch/peterfleury/avrgcc_411_windows.html

-Ripa

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ripa,

Has your code burst through 64KB yet?

Cliff

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Aha!
No, it hasn't. So when 32-bit addresses are really needed, they appear in the .elf? Now I'm having 16-bits like before?

-Ripa

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

> No, it hasn't. So when 32-bit addresses are really needed,
> they appear in the .elf?

No. This is an on/off switch to be taken at the compiler's
compile time. Either the generated DWARF-2 information uses
16-bit pointers, or it uses 32-bit pointers. Eric pulled the
switch in some WinAVR version once (I think in January, 2006),
and the generated 32-bit DWARF-2 pointer information made AVR
Studio's ELF/DWARF-2 parser crash. So he quickly reverted the
change.

Supposedly, according to the information recorded in the respective
GCC bug entry, AVR Studio is now handle to automatically detect the
pointer size of the debug information. However, rumours are that
this particular version has not yet been published.

> Now I'm having 16-bits like before?

Yes, you are.

Jörg Wunsch

Please don't send me PMs, use email if you want to approach me personally.