rje 2002/12/09 04:21:45 PST
Modified files: (Branch: speculative-evaluation)
ghc/compiler/codeGen CgExpr.lhs
ghc/compiler/main CmdLineOpts.lhs
ghc/compiler/profiling SCCfinal.lhs
ghc/compiler/simplStg SRT.lhs
ghc/compiler/stgSyn CoreToStg.lhs StgLint.lhs StgSyn.lhs
Log:
Early stages of the rewrite of my speculation code, based on the current GHC tree.
(in a new branch, all of its own)
This version attempts to be a bit less of a hack than the original version, with the
hope that it might reach a stage where it can be merged into normal GHC.
In particular, I now have a new "StgSpecLet" constructor in StgCode, rather than
just adding an extra boolean flag to StgCase to say that a case is actually a
speculative let in disguise (and thus something that the RTS can alter at runtime).
This commit is a bit of a mess, containing some bits of code that I'm fairly sure
are wrong, but it's a start, and gives other people something to look at and pick
holes in.
Revision Changes Path
1.51.2.1.2.1 +11 -1 fptools/ghc/compiler/codeGen/CgExpr.lhs
1.169.4.1 +3 -0 fptools/ghc/compiler/main/CmdLineOpts.lhs
1.35.10.1 +3 -0 fptools/ghc/compiler/profiling/SCCfinal.lhs
1.16.16.1 +19 -0 fptools/ghc/compiler/simplStg/SRT.lhs
1.96.2.1.2.1 +76 -9 fptools/ghc/compiler/stgSyn/CoreToStg.lhs
1.41.4.1 +9 -0 fptools/ghc/compiler/stgSyn/StgLint.lhs
1.38.14.1 +30 -0 fptools/ghc/compiler/stgSyn/StgSyn.lhs
_______________________________________________
Cvs-ghc mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/cvs-ghc
