[rfc] slightly better reload constant rematerialization

[Richard Henderson]

On Thu, Jan 20, 2005 at 03:29:00PM -0500, Ian Lance Taylor wrote:
> Also, the current cost metric doesn't let us compare sequences of
> instructions.  For example, we may happen to know that we can do an
> add and a multiply in a single cycle, but that two adds or two
> multiplies take two cycles.  How do we express that kind of thing
> using the cost metrics we have today?

We compute the cost of (plus (mult x y) z).


r~