This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
RE: [Gc] Re: boehm-gc merge for GCC
- From: "Boehm, Hans" <hans dot boehm at hp dot com>
- To: Bryce McKinlay <bmckinlay at gmail dot com>, GCC Java <java at gcc dot gnu dot org>
- Cc: Ivan Maidanski <ivmai at mail dot ru>
- Date: Tue, 27 Nov 2012 01:08:20 +0000
- Subject: RE: [Gc] Re: boehm-gc merge for GCC
- References: <1353097280.505328435@f161.mail.ru> <50A67667.7020600@ubuntu.com> <CAF1jjLucRudE2a3gCZq6erkJKfs1nC+DicM+4f4moXJ5HPy9g@mail.gmail.com> <1353099257.140499471@f123.mail.ru> <CAF1jjLsA1nzLaJejMiAfzqZZ8btw37n0B3H7_qz-UrLvNQGqzA@mail.gmail.com> <50A6EE27.4000605@ubuntu.com> <50ABADE0.9000305@redhat.com> <50B385C5.6010004@ubuntu.com> <CALUNu-p3muXCwGV9sRSF550EvyM2=f0=mbEzDEhtf=BdhUJKMQ@mail.gmail.com> <CALUNu-pkdM8xXrJoCuJTW41zn0BiHntQLzC8=sLuHR2DbHVjbw@mail.gmail.com>
>
> > - bdwgc now depends on another library (libatomic-ops). Shouldn't GCC
> > be able to use it's atomic builtins for some supported targets at
> > least?
>
>
> Yes, but it sounds like far more effort than its worth to go and change
> it
> just for the sake of using builtins. Maybe libatomic-ops itself could
> be
> made to use the builtins when available.
>
Libatomic_ops should eventually be configurable to just use C11/C++11 atomic operations. That hasn't happened yet, but should be fairly easy. Or, if we wait long enough, we could just use the C11 primitves directly. Historically libatomic-ops preceded and influenced the C11/C++11 interface, but the C11/C++11 interface is a much better design (another iteration, a lot more careful thought by many more people, etc.)
Hans