This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: g++ and aliasing bools
- From: Daniel Berlin <dan at dberlin dot org>
- To: Paolo Carlini <pcarlini at unitus dot it>
- Cc: Gabriel Dos Reis <gdr at codesourcery dot com>, <gcc at gcc dot gnu dot org>
- Date: Fri, 25 Jan 2002 07:42:34 -0500 (EST)
- Subject: Re: g++ and aliasing bools
On Fri, 25 Jan 2002, Paolo Carlini wrote:
> Gabriel Dos Reis wrote:
>
> >hey, isn't Mark who
> >implemented the type-based analysis in g++? ;-)
> >
> I know that, Gaby.
>
> And I already complimented him for his beautiful article in Dr Dobb's
> where he described the whole issue! (it never occurred to me before that
> C-type languages could be optimized better than fortran!)
>
> ... however, I learned a few months ago from some of Kenner's postings
> (and my personal experimentation) that right now the C++ alias analysis
> *actually* implemented in g++ is not so strong after all... :-(
In fact, it's turned off for structs, classes, and, well, any aggregate
type.
This is the problem.
--Dan