USE patch vs the PA

[Jeffrey A Law]

  In message < 199710250200.WAA00802@jfc. >you write:
  > 
  > > There were basically three problems with the alias code.  I'll write
  > > them up in more detail later this weekend.
  > > 
  > >   * Didn't handle autoincrement addressing.  It just punted.
  > >   autoincrement addresses can be treated just like a set of the
  > >   base value to itself plus a small constant.
  > 
  > I thought it would work with autoincrement addressing.  It uses
  > note_stores to detect sets, and note_stores only looks for explicit
  > sets.
It's not a major problem -- find_base_term and find_base_value give up
when they encounter an autoincrement address.

Stripping off the autoinc and exposing the base register should be all
we need to do unless I'm missing something important (which is certainly
possible!)


  > >   * find_base_value (psuedo) would return the pseudo instead of
  > >   the base value of the pseudo.
  > 
  > That's how it was intended to work.  Insns are not scanned in the same
  > order they are executed.  Consider insns which appear in rtl in this
  > order:
You're absolutely right -- I missed the propagation code at the bottom
of the file which should have figured this out.

jeff