This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: [boehms-gc] Performance results
- From: Daniel Berlin <dberlin at dberlin dot org>
- To: Laurynas Biveinis <laurynas dot biveinis at gmail dot com>
- Cc: gcc at gcc dot gnu dot org
- Date: Tue, 25 Jul 2006 07:43:14 -0400
- Subject: Re: [boehms-gc] Performance results
- References: <d1e3ff2b0607240145h1558bc59h61a6250150b14fd7@mail.gmail.com>
>
> 6) Are there any data objects in GCC, that the only pointers pointing
> to them point to their interiors and not the beginning? If I could
> disable Boehm's interior pointer support, that should boost
> performance in all respects.
There probably are.
But at some point, if you are tuning boehm's to this point, what is the
advantage of it over just writing your own collector like ggc-page? :)
I mean, you might as well drive boehm's marking using the exact marking
process (IE replacing the ggc_mark calls with GC_mark_*, instead of it's
conservative one, you'd probably get better performance since it
wouldn't be guessing :)
Also, I don't think we have any large atomic structures, TBQH.
--Dan