Flags parameter in comptypes

[Jim Wilson]

Joseph S. Myers wrote:
> Thus the only point I can see in the parameter is that it makes the
> interface compatible with the C++ front end - but this isn't being used
> anywhere (no common code calls comptypes).  Do we really want to keep an
> unused parameter passed among four functions on the basis that it might be
> useful for interface compatibility in future?

I'd say no.

> Would it make sense to file a feature-request PR asking that
> -Wunused-parameter detect such situations?

That would require intraprocedural analysis and would be hard to do on 
RTL, which is where the current checks are done.  I don't think it would 
be feasible to add this without completely rewriting the code to operate 
on trees or gimple.  I don't think this would be a very useful PR.  That 
is just my opinion though.
-- 
Jim Wilson, GNU Tools Support, http://www.SpecifixInc.com