Wrong RTL instruction deleted

[Andrew Pinski]

On Dec 27, 2004, at 9:03 AM, Robert Dewar wrote:

> Presumably the Ada program would fail in the same way. I have
> felt for a while that we should probably set -fno-strict-aliasing
> by default in GNAT, and if this is really true, it adds fuel to
> the argument for doing this.


If the Ada front-end's get_alias_set is correct with this respect
there should be no problem and -fstrict-aliasing should just work.
Otherwise I and others would consider this to be a bug in the Ada
front-end.

Thanks,
Andrew Pinski