This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: moving constants out of loops



  In message <9806302349.AA14600@vlsi1.ultra.nyu.edu>you write:
  > I don't see it as an alias issue per se.
Certainly the alias issue is a secondary effect -- but a very important
one -- on the PA at least the lossage due to poor alias analysis
outweighed all the benefits we originally got from finding more givs.

  > What I see is that the more
  > givs that are found, the more relationships between givs that can be
  > found.
Yup.

  > Currently loop.c doesn't take advantage of that.  Part of what
  > I was doing (and never finished) is to do much better tracking of the
  > relationship between givs.
This is one of the key things rth has been working on.

  > That allows for much better choosing of
  > which to reduce and also to allow many more cases of biv elimination.
Yup.

  > I think this is precisely the trickiest issue in all of loop.c.
Richard has found that out the hard way.  I think we're up to the
13th or so public revision.  I think his stuff is real close to being
ready.

jeff


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]