-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi,
On 05/31/11 18:24, Tobias Burnus wrote:
> This patch adds the atomic_define and atomic_ref intrinsics. They are
> currently implemented in form a simple assignment. For -fcoarray=lib
> they will be replaced by a function call to libcaf_{single,mpi,*}.
>
> I was shortly thinking of using something more clever, but I concluded
> that a simple assignment should be sufficiently atomic (for
> -fcoarray=single).
>
> Build and regtested on x86-64-linux.
> OK for the trunk?
Ok. Just one thought:
+ if (!(atom->ts.type == BT_INTEGER && atom->ts.kind == gfc_c_int_kind)
+ && !(atom->ts.type == BT_LOGICAL && atom->ts.kind == gfc_c_int_kind))
What about defining another constant for atomic_int_kind rather than
gfc_c_int_kind, in case we want to change this in the future? Or is it
clear that we indeed *always* want atomic_int_kind to be c_int_kind?
Yours,
Daniel
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.1 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/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=fYF2
-----END PGP SIGNATURE-----
