This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [gomp] RFC: GOMP_CPU_AFFINITY
- From: Richard Henderson <rth at gcc dot gnu dot org>
- To: Jakub Jelinek <jakub at redhat dot com>
- Cc: Diego Novillo <dnovillo at redhat dot com>, gcc-patches at gcc dot gnu dot org
- Date: Mon, 22 May 2006 10:54:12 -0700
- Subject: Re: [gomp] RFC: GOMP_CPU_AFFINITY
- References: <20060519153701.GL13709@devserv.devel.redhat.com>
On Fri, May 19, 2006 at 11:37:01AM -0400, Jakub Jelinek wrote:
> The patch as is immediately binds the initial thread to the first CPU from
> the list if the env var is set, an alternative perhaps could be to only do
> that when the initial thread enters a parallel region and reset it to the
> old affinity mask when it leaves it, what do you think?
Nah, what you have is fine.
I'm not sure I see the real utility in it, since I don't
really believe the Sun claim that you'll see threads
switching processors often. But I suppose it doesn't
really add any overhead to go ahead and provide the
functionality.
r~