PATCH: VAX sltu and sgeu patterns

[Richard Henderson]

On Mon, Aug 21, 2000 at 11:57:22AM -0400, John David Anglin wrote:
> I don't understand why reload couldn't be fixed to push the reload back
> before the cc0 setter if required.  This might be a lot less work.

Actually, that's pretty hard, since now you have to figure
reload registers not conflicting with some other instruction.
The easiest thing to do overall is to pretend the setter and
user are one instruction until after reload.


r~