Please, RedHat and/or Cygnus people on this list...

[Eric Christopher]

> What matters to me is what can be done to mitigate the problems this
> decision has caused.
> 
> Here is my personal situation (reading the mailing lists archives
> it seems others are having similar problems).


To be honest, I've had less problems with this compiler than I did with
the 2.95.x series, but YMMV.


> 
> The compiler shipped with RH7 cannot compile the C++ application
> on which my colleagues and myself are working.
> One of the problems has been filed using the GNAT system (#687).

You should probably report this to bugzilla.redhat.com, this is where
bugs against Red Hat releases should be filed and requests of this sort
should be made.

This problem is fixed in current cvs sources (I tried with -O0 through
-O3), though I did notice a segfault without -fno-operator-names.  While
I believe the code would be incorrect then, the compiler still shouldn't
seg fault.  I'll add notes to the gnats entry about that.

-eric