[Haskell-cafe] GHC's parallel garbage collector -- what am I doing wrong?

[Michael Lesniak]

Hello Bryan,

> The parallel GC currently doesn't behave well with concurrent programs that
> uses multiple capabilities (aka OS threads), and the behaviour you see is
> the known symptom of this.. I believe that Simon Marlow has some fixes in
> hand that may go into 6.12.2.


> Are you saying that you see two different classes of undesirable
> performance, one with -qg and one without?
No, I wouldn't say that.


> How are your threads in your real
> program communicating with each other? We've seen problems there when
> there's a lot of contention for e.g. IORefs among thousands of threads.
No, my program uses only MVars and Chan and much less threads,
normally as many threads as cores available (doing numbercrunching).


But, as Daniel pointed out, I believe it has something to do with
Ubuntu's kernelpatches. I tried another machine with an earlier kernel
(2.6.28...) and the same program did not behave as bad.

Currently I'm trying some experiements with custom kernels and if I
solve this using a compiled kernel I'll blog about this.


Cheers,
  Michael