#14257: Heap profiling with ghc and hp2ps and strict function application ($!) gives samples out of sequence (regression) -------------------------------------+------------------------------------- Reporter: carlostome | Owner: bgamari Type: bug | Status: new Priority: high | Milestone: 8.4.1 Component: Runtime System | Version: 8.2.1 Resolution: | Keywords: profiler Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: #14006, #11645 | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by bgamari): Thanks for picking this up, duog! I was in the middle of writing a reply starting with,
The issue with this approach is that we will now have runtime which won't be accounted for. The heap census can be a considerable amount of time... when I noticed that we already have a separate set of fields for accounting heap census time. Presumably this means that we were double- counting time previously. Anyways, this suggests to me at least that your approach should be fine.
-- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/14257#comment:9 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler