#7847: Maintain per-generation lists of weak pointers ---------------------------------+------------------------------------------ Reporter: akio | Owner: Type: feature request | Status: infoneeded Priority: normal | Milestone: Component: Runtime System | Version: 7.7 Keywords: | Os: Unknown/Multiple Architecture: Unknown/Multiple | Failure: Compile-time performance bug Difficulty: Unknown | Testcase: Blockedby: | Blocking: Related: | ---------------------------------+------------------------------------------ Comment(by akio): Thank you for review. Replying to [comment:9 simonmar]:
patch 0001 looks good to me.
On patch 0002:
* Why do we have `dead_weak_ptr_list_tail` rather than just consing new items on the front of the list?
This is a trace of an earlier design that survived by accident. I'll clean it up in a few days.
If you haven't already, I recommend testing these patches with
`libraries/base/tests/memo001.hs` and `libraries/base/tests/memo002.hs`. Those are pretty good stress tests of the weak pointer implementation. Yes, the patches passed these tests (and the rest of validation). -- Ticket URL: http://hackage.haskell.org/trac/ghc/ticket/7847#comment:10 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler