This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [tree-ssa] Enhance statement checking & fix minor buglets
On Mon, 2004-04-05 at 19:12, Daniel Berlin wrote:
> On Apr 5, 2004, at 6:42 PM, Richard Henderson wrote:
>
> > On Mon, Apr 05, 2004 at 02:12:46PM -0400, Diego Novillo wrote:
> >> Well, one thing that occurred to me would be to re-associate these
> >> expressions to be PLUS_EXPR<SSA_NAME,MINUS_EXPR>. That would be a
> >> valid
> >> gimple expression if MINUS_EXPR is_gimple_min_invariant. Though that
> >> strikes me as somewhat hackish.
>
> > True, but it might be good enough. I'm working on it.
> >
> This would then be the only binary expression that allowed a binary
> expression as one of it's operands, which would seriously defeat the
> purpose of GIMPLE.
>
That's precisely what I don't like about the approach. Putting special
cases in the IL is not a good thing.
How about the opposite? Why don't we dissallow ADDR_EXPR+/-CONST_INT
when we determine that it cannot be folded into array syntax?
Diego.