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 bgamari): Replying to [comment:21 rwbarton]: performance improvement from non-branching code? I must say I'm still rather skeptical of this whole project. You raise some fair points and perhaps Joachim's findings suggest that it it's not worth bending over backwards to accommodate branch-free code. Looking further into the performance implications of these branches is high on my to-do list. Regardless, there may be value in allowing rules to examine the application context even beyond this particular optimization (although at the risk of making RULEs a bit less predictable than they are currently). -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/9661#comment:22 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler