alpha himode reload problem

[Richard Henderson]

On Sun, Oct 26, 1997 at 04:09:47PM -0800, Richard Henderson wrote:
> Comments on the following?  The recog.c changes were necesary to 
> get combine not to undo our hard work in stmt.c.

Blah.  I should have tested this on something real like the kernel.
It needs a bit more thought, but comments on what ought to happen
are still welcome.

In the meantime, the attached patch fills out the brute force solution
Jim posted, for 8 cases (also attached) I could think of.  Any other
points that are likely to be called?

At least the kernel builds now.  We'll see how well it boots in a 
moment.  ;-)


r~