RE: [avr-gcc-list] Latest working versions (ATmega324p)?

[Weddington, Eric]

> -----Original Message-----
> From: Rick Mann [mailto:address@hidden 
> Sent: Thursday, April 24, 2008 12:08 PM
> To: Bernard Fouché
> Cc: Weddington, Eric; avr-gcc List
> Subject: Re: [avr-gcc-list] Latest working versions (ATmega324p)?
> 
> 
> On Apr 24, 2008, at 3:55 AM, Bernard Fouché wrote:
> 
> > Currently if one does not read all the traffic on this 
> mailing list,  
> > it's difficult to know what patches are required to have a fully  
> > fonctionnal avr-gcc toolchain.
> 
> +1
> 
> > I'd like to upgrade my toolchain on Linux, and if I read the list  
> > correctly, the surest solution is first to get WinAvr 
> 20080411, grab  
> > the included patches and apply them to stock FSF source code to be  
> > able to have, at last, a working Linux toolchain. (Linux needing  
> > files first released for Windows... arg!)
> 
> I'm on a Mac, and don't know how to unpack the .exe file.
> 
> > Could the exact list of files/patches to use be listed 
> somewhere on http://www.nongnu.org/avr-libc 
> >  website?
> 
> +1
> 
> > Or will the release of 4.3.1 fix everything?
> 
> 
> There will always be a need for listing patches. It took me 
> forever to  
> figure out just one patch file, with its many many CVS revisions.

This is why I have started to post the patches on the WinAVR CVS repository, 
tagged with the relevant release number. The patches are already ordered, for 
your convenience, so you don't have to figure out which to one to patch first. 
The patches are also consistent, in that they are all patched using --strip=0 
from the top level of the source directory. They are categorized so you don't 
have to figure out which ones are host-specific (mingw) so you don't have to 
apply those to a Linux-hosted toolchain.

I typically post the patches I use in WinAVR *after* a WinAVR release. You will 
notice that I have yet to post the patches that I use for 20080411. If you look 
carefully, you will also notice that I have pulled the 20080411 release, and it 
is because of a wrong code generation bug. I couldn't, in good conscience, 
leave that release up there knowing that it generated bad code.

I'm currently working on a new WinAVR, with the offending patch removed, and I 
hope that this will finally clear up the problems. After I release WinAVR, and 
no one is screaming that it is causing bad code to be generated then I will 
post the patches on the WinAVR CVS.

Sorry for all the hassle.

Eric Weddington