This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Speedup iterative_hash_expr
> On Thu, Sep 02, 2004 at 08:41:54PM +0200, Jan Hubicka wrote:
> > + /* Produce good hash value combining VAL and VAL2. */
> > + static inline hashval_t
> > + #ifdef __STDC__
> > + iterative_hash_hashval_t (hashval_t val, hashval_t val2)
> > + #else
> > + iterative_hash_hashval_t (val, val2)
> > + hashval_t val, val2;
> > + #endif
>
> I think you should not provide these inline unless "inline"
> is actually supported, meaning GCC or C++.
>
> > + #ifdef HOST_WIDE_INT
>
> I *really* do not like this. libiberty should not be
> referring to HWI at all.
Well, me neither but...
>
> Why don't you just put these in tree.c with the other
> iterative hash bits and be done with it?
... it seemed that it might be usefull in other cases too. For example
CSE cannon_hash needs similar tricks but uses + for merging instead.
I am not sure if those sophisficated merging functions actually makes
any difference tought and I will be happy to put it in tree.c as the
hashtable.h way is ineed particularly ugly.
Honza
>
>
>
> r~