Re: Why alias set 0 ?

[Paolo Carlini <pcarlini at unitus dot it>]

Hi,

Richard Kenner wrote:

>     I am just looking at the attached testcase and first thing that
looks
>     odd to me is that there are memory references of alias set zero to
the
>     'n' field of RealMatrix structure:
>
>     2.95 version appears to have similar problem, but why?
>
>     Richard, do you have any idea?
>
> I said why a few days ago:  It's specific to C++ and is in:
>
> static HOST_WIDE_INT
> cxx_get_alias_set (t)
>      tree t;
> {
>   /* It's not yet safe to use alias sets for classes in C++ because
>      the TYPE_FIELDs list for a class doesn't mention base classes.
*/
>   if (AGGREGATE_TYPE_P (t))
>     return 0;
>
>   return c_common_get_alias_set (t);
> }

First thank you very much Jan and Richard for your attention to my
personal battle against "Haney speed" :-)
Anyway, at this point, I think I have to mention two different things:
1- Richard mentioned in fact that possible problem with alias set zero a
few days ago:

    http://gcc.gnu.org/ml/gcc/2001-10/msg01398.html

Sorry Jan if I did'nt told you about that. At the time, howewer, Richard
did'nt point to specific line in the sources.

2- Besides any interesting problem with the C++ aliasing treatment,
which I would like to understand but I can't, unfortunately, right now,
my question is: *why* no more than 10 days ago (in some utterly
misterious way) the compiler managed to deal much better with that
testcase????

Thanks,
Paolo.