#8272: testing if SpLim=$rbp and Sp=$rsp changed performance at all -------------------------------------+------------------------------------- Reporter: carter | Owner: carter Type: task | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.7 Resolution: | Keywords: | callingConvention Operating System: Unknown/Multiple | Architecture: x86_64 Type of failure: Runtime | (amd64) performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by simonmar): @bgamari: doesn't the DWARF stack unwinding support that you've been working on help with `perf`? @michalt: I think we have to do (1) in order to do call/ret, because otherwise the stack would be split over two places, and the RTS would have a terrible time walking it. Or perhaps I've misunderstood what you mean? I'm actually not all that enthusiastic about the proposal having just re- read https://github.com/ghc-proposals/ghc-proposals/pull/17. The benefits are small or are actually regressions (code size and the overhead for jmp after call) and it's a huge upheaval. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8272#comment:16 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler