Mingw Tools 4.7

[Paul Whitfield]

Ralf Corsepius wrote:
> On Wed, 2008-04-16 at 11:57 +0800, Paul Whitfield wrote:
>> Hi All,
>>
>> I am trying (in a spare minute) to drag my application out of the RTEMS
>> dark ages onto a current RTEMS version. Currently I am stuck at rtems 
>> 4.6.6  for Coldfire on Windows :-(
> I would recommend you to switch to using Linux, but I guess this is not
> an option to you, otherwise you would not already be using MinGW :)

Indeed :-(.

>> I have tried to go to 4.8, how-ever the current gcc in the rtems4-8
>> for mingw gets ICE's for Rtems and my code so that is not an option.
> Coldfire targets? ... Are you sure this issue is related to MinGW?

No, you are quite right! It is more likely a compiler issue

I should not assume that it is all Windows fault.. but usually it is a 
safe bet :-)

> 
> There have been quite a number of reports related to "gcc-4.2.x ICE'ing
> for m68k targets" on all kind of build-hosts, which makes it very likely
> for your issues not to be related to MinGW at all.
> 
> Could you provide further details about the ICE you are observing?

I have reported it via the rtems Bugzilla.

https://www.rtems.org/bugzilla/show_bug.cgi?id=1281

Let me know if you need any more details.


>> So I tried to go for 4.7. Downloading and installing 4.7-4 for Mingw
>> from the rtems server actually installs a 4.8 tool chain!
> Chris would be the person to answer this. I don't know if he is
> listening, because he recently mentioned to be away for some weeks.

Yes I though he would be the one to talk to !

Looking at the package versions the 4.7-3 tools are not quite
up to date with the information on the rtems site for 4.7.2

Newlib is 1.14.0-5 instead of 1.15
Binutils is 2.16.91 instead of 2.17
automake is 1.09 instead of 1.10

rtems configures and builds OK, so I guess that automake is not an issue.

Sigh..

Regards

Paul