lno-branch vs ptr-to-member

[Andrew Pinski]

> > Calling convert from the middle-end is asking for troubles.
> 
> interesting; we do it moreorless everwhere (dom, sra, ccp, ...).

That is because they have not been changed yet as fold_convert used to
be only local to fold-const.c.  I am testing a patch to change them into
fold_convert right now.

> > One it creates non-gimple,
> 
> Why should this be a problem (as long as you know about it)?

True but there was an use in phiopt which I changed because the use
did not take that into account.

> > two it prints out error message which is bad for the middle.
> 
> Yes, this seems worse.
> 
> > I posted a patch to change all of the convert to fold_convert in tree-scalar-evolution.c
> > at <http://gcc.gnu.org/ml/gcc-patches/2004-06/msg00460.html>.
> 
> Feel free to install the patch.  Thanks.
> 
> Zdenek


Thanks,
Andrew Pinski