[PATCH] Complete useless_type_conversion_p_1 some more (need Ada help)

[Richard Guenther]

On Tue, 4 Aug 2009, Jakub Jelinek wrote:

> On Tue, Aug 04, 2009 at 01:46:20PM +0200, Richard Guenther wrote:
> > 
> > This completes useless_type_conversion_p_1 some more by making the
> > comparison for function and array types be independent of frontends.
> > 
> > During the array work I came along some oddities wrt the Ada frontend
> > and the stuff that is left in the types.  For example if you remove
> > the hack
> > 
> > +       /* ???  Ada does weird things with type sizes and defers 
> > gimplifying
> > +          them.  To not ICE during expansion for for example cdd2a01 we
> > +        avoid stripping conversions from types whose size contains
> > +        placeholder exprs.  */
> > +       if (CONTAINS_PLACEHOLDER_P (TYPE_SIZE (inner_type)))
> > +       return false;
> 
> Couldn't we have a global variable which says if there are any
> types containing placeholders (i.e. whether any PLACEHOLDER_EXPRs have been
> created)?  That would make this cheaper for all the languages that don't have
> them (all but Ada) and could be used even for LTO FE.

If this (and similar places) turn out to be expensive we probably could
spare a (or two) bit(s) in the type tree to cache this?

I'm experimenting with checking TYPE_SIZE for being constant instead now.

Richard.