Did anyone ever come up with actual benchmarks that show a potential
#9661: Branchless ==# is compiled to branchy code -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: bgamari Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.9 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Type of failure: Runtime | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Revisions: D854 -------------------------------------+------------------------------------- Comment (by jstolarek): Replying to [comment:21 rwbarton]: performance improvement from non-branching code? https://ghc.haskell.org/trac/ghc/wiki/PrimBool#Benchmarks I know these numbers are not much, but they give some hope that branchless operations can improve performance. I recall that Gregory Collins saw a potential use for branchless primops in his `hastables` library. When I worked on implementing #6135 I did some measurements on hastbales but saw now improvement in performance - now I realize that this might have been because `==#` was compiled to branchy code but I didn't realize that. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/9661#comment:24 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler