#1600: Optimisation: CPR the results of IO -------------------------------------+------------------------------------- Reporter: simonmar | Owner: Type: task | Status: new Priority: lowest | Milestone: Component: Compiler | Version: 6.6.1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Type of failure: Runtime | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: #8598 | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by akio): I've rebased nomeata's branch: https://github.com/takano- akio/ghc/compare/afdde48...f8eb150 Some nofib results: {{{ -------------------------------------------------------------------------------- Program Size Allocs Runtime Elapsed TotalMem -------------------------------------------------------------------------------- mandel +1.9% -25.0% 0.047 0.047 0.0% gcd +2.0% -21.6% 0.030 0.031 0.0% k-nucleotide +1.7% -18.0% +3.4% +2.7% 0.0% wave4main +2.0% +12.3% 0.179 0.179 0.0% kahan +2.0% +7.8% +1.4% +1.3% 0.0% solid +2.0% -6.4% 0.081 0.081 0.0% sphere +2.0% +5.8% 0.040 0.040 0.0% hpg +2.0% +4.1% 0.102 0.102 0.0% symalg +1.9% +3.6% 0.007 0.007 0.0% gamteb +2.4% -4.8% 0.026 0.026 0.0% -------------------------------------------------------------------------------- Min +1.5% -25.0% -0.6% -0.4% -20.0% Max +2.8% +12.3% +12.7% +12.8% +12.0% Geometric Mean +2.0% -0.5% +5.0% +4.7% +0.2% }}} I'll look into the allocation regressions next. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/1600#comment:52 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler