Re: [GHC] #5344: CSE should look through coercions

#5344: CSE should look through coercions -------------------------------------+------------------------------------- Reporter: reinerp | Owner: simonpj Type: feature request | Status: closed Priority: normal | Milestone: ⊥ Component: Compiler | Version: 7.0.3 Resolution: duplicate | Keywords: cse 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: | -------------------------------------+------------------------------------- Comment (by rwbarton): Hah, you confused me by writing this comment before updating #9291 :) This example seems easier than the original example in #9291 though, since `f` can be expressed using a cast as a well-typed program in the current Core language. All else being equal, it seems better to do the CSE earlier rather than later, though I'm not sure whether it makes an actual difference in practice. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/5344#comment:5 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
participants (1)
-
GHC