#13002: :set -O does not work in .ghci file -------------------------------------+------------------------------------- Reporter: George | Owner: (none) Type: bug | Status: closed Priority: normal | Milestone: 8.0.3 Component: GHCi | Version: 8.0.1 Resolution: worksforme | Keywords: | RecompilationCheck Operating System: Unknown/Multiple | Architecture: Type of failure: Runtime | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Changes (by dfeuer): * status: new => closed * resolution: => worksforme * milestone: 8.4.1 => 8.0.3 Comment: I'm not at all convinced that this bug is valid at all. I suspect the timing difference you saw was some sort of insignificant artifact; perhaps the extra command line options are affecting GC timings. I tested with a version that includes an utterly bogus `RULES` pragma: {{{#!hs {-# OPTIONS_GHC -Wall #-} module T13002 where testFromTo :: Int -> Int testFromTo n = length ([0..(10^n)] :: [Int]) {-# NOINLINE testFromTo #-} test :: Int -> Int test x = testFromTo x {-# RULES "bogus" forall x. testFromTo x = 12 #-} }}} This produces semantically different `test` functions depending on whether optimizations are enabled or not, and the change occurs as expected when I add `-O` to the `.ghci` file. It works for me under 8.0, 8.2, and HEAD. So I'm going to close this for now; if you disagree, I suggest you add a more robust test case. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13002#comment:8 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler