Hoopl vs LLVM?
michal.terepeta at gmail.com
Wed Dec 12 18:28:19 CET 2012
On 12.12 09:06, Greg Fitzgerald wrote:
> On Wed, Dec 12, 2012 at 4:35 AM, Simon Marlow <marlowsd at gmail.com> wrote:
> > Now, all that LLVM knows is that z was read from Sp, it has no more
> > information about its value.
> Are you saying Hoopl can deduce the original form from the CPS-version? Or
> that LLVM can't encode the original form? Or within GHC, LLVM is thrown in
> late in the game, where neither Hoopl nor LLVM can be of much use.
I think what happens is that in GHC the Hoopl optimizations are
performed before the CPS transformation whereas LLVM runs after the
transformation. So Hoopl will have the access to the easier version of
the above code and LLVM to the more difficult one.
More information about the Glasgow-haskell-users