This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH][RFC] A GIMPLE and "GIMPLE type system" verifier


On Fri, 1 Jun 2007, Richard Kenner wrote:

> > The other option i proposed was to add a new field called
> > TYPE_EQUIVALENCE, representing the equivalence set of a type.  The
> > gimpliifer would simply unify all types using some equivalence metric
> > (structural, name, whatever), applied uniformly to all types.  Two
> > types are equivalent if they have the same TYPE_EQUIVALENCE.
> 
> What I don't understand is why the middle-end or anything past it would care
> about the front-end view of what types are equivalent.  Isn't it too late to
> matter?
> 
> Why isn't the only thing that matters to GIMPLE whether two types will
> generate the same RTL?

It is (should be) that way.  If you disable calling into the frontend
in tree_ssa_useless_type_conversion_1 you should only get some more
explicit conversions and a few testsuite regressions due to missed
optimizations.  At least the last time I tried.

Richard.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]