Repository : ssh://darcs.haskell.org//srv/darcs/ghc On branch : supercompiler
http://hackage.haskell.org/trac/ghc/changeset/fa5cfc059be1376b91b39253e606c7973cc821e9 >--------------------------------------------------------------- commit fa5cfc059be1376b91b39253e606c7973cc821e9 Author: Max Bolingbroke <batterseapo...@hotmail.com> Date: Thu Jan 5 09:26:45 2012 +0000 Comment only >--------------------------------------------------------------- compiler/supercompile/Supercompile/Drive/Match.hs | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/compiler/supercompile/Supercompile/Drive/Match.hs b/compiler/supercompile/Supercompile/Drive/Match.hs index 4ad4317..a28a27b 100644 --- a/compiler/supercompile/Supercompile/Drive/Match.hs +++ b/compiler/supercompile/Supercompile/Drive/Match.hs @@ -393,6 +393,11 @@ matchECFrame init_rn2 rn2 kf_l kf_r = go (tagee kf_l) (tagee kf_r) -- a |-> True; ()<(a, a)> `match` c |-> True; d |-> True; ()<(c, d)> -- a |-> True; ()<(a, a)> `match` c |-> True; d |-> c; ()<(c, d)> -- However, I'm going to reject this for now (simpler). +-- +-- TODO: arguably I should check that this is actually a true *bijection* not just a *function* because a renaming like +-- {x |-> a, y |-> a} means that if we carrried on supercompiling here we could exploit more equalities (via positive information +-- propagation - imagine we scrutinise x and later scrutinise y) and potentially get better code that at the tieback site. I need to +-- check how important this is in practice. safeMkMatchResult :: [(Var, Var)] -> Match MatchResult safeMkMatchResult eqs = guard "safeMkRenaming" (all (\(x_l, x_r) -> M.lookup x_l eqs_map == Just x_r) eqs) >> return eqs_map where eqs_map = M.fromList eqs _______________________________________________ Cvs-ghc mailing list Cvs-ghc@haskell.org http://www.haskell.org/mailman/listinfo/cvs-ghc
