Question on scalar replacement

[Richard Henderson]

On Mon, Jun 14, 2004 at 10:17:34AM -0400, Diego Novillo wrote:
> Jason, is a.f = foo() valid GIMPLE?  If not, then the bug is in the
> gimplifier.  Otherwise, the patch I sent should be OK.

We can do better than your patch.  If is_sra_candidate_ref, we
can scalarize 

	s.f = foo()

to 

	SR.105 = foo()

rather than

	s.f = foo()
	SR.100 = s.a;
	SR.101 = s.b;
	SR.102 = s.c;
	SR.103 = s.d;
	SR.104 = s.e;
	SR.105 = s.f;


r~