Re: combine_givs_p address cost

[Jeffrey A Law <law at cygnus dot com>]

  In message <19981123161130.A10415@dot.cygnus.com>you write:
  > On Tue, Nov 24, 1998 at 11:04:35AM +1300, Michael Hayes wrote:
  > > In my example, combine_givs concludes that the GIV with the highest
  > > benefit is in the insn (25) that explicitly multiples a BIV by a
  > > constant.  However, this is a poor choice for GIV combination since
  > > this insn can usually be optimised away.
  > 
  > Oh, I see now.  Your problem is with those occasionally irritating
  > intermediate values.
  > 
  > Here's a possible approach.  It happens to work for your testcase;
  > I would be interested to find out what happens on other tests.
  > 
  > Toon, I'd like for you to be able to try this out as well.  I didn't
  > know you were still having problems with the mainline though.  I 
  > think Jeff has an old m68k somewhere -- I'll see about trying a build
  > and see if I can find the problem.
Actually, I don't have a m68k anymore.  The UofU needed it back.

Worse, it appears as if the last 2 or 3 m68k boxes at the UofU that I knew
about have either been turned off or converted into PAs.

Anyone out there have a m68k that they might be able to grant remote access
to so that we can track down Toon's problem?

jeff