This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [RFC PATCH] Optimized __cxa_guard_{acquire,release,abort} for Linux
- From: Jakub Jelinek <jakub at redhat dot com>
- To: Benjamin Kosnik <bkoz at redhat dot com>
- Cc: Jason Merrill <jason at redhat dot com>, gcc-patches at gcc dot gnu dot org, libstdc++ at gcc dot gnu dot org
- Date: Sun, 6 Jan 2008 15:35:17 -0500
- Subject: Re: [RFC PATCH] Optimized __cxa_guard_{acquire,release,abort} for Linux
- References: <20071231133948.GQ20451@devserv.devel.redhat.com> <20080104165012.7aff650f@wabash.artheist.org>
- Reply-to: Jakub Jelinek <jakub at redhat dot com>
On Fri, Jan 04, 2008 at 04:50:12PM -0600, Benjamin Kosnik wrote:
>
> > The patch below doesn't include the configury bits yet (they might be
> > very similar to enable_linux_futex stuff in libgomp configury).
> > Would this be acceptable with the configury bits in even for 4.3
> > (could be e.g. only enabled if explicitly requested with
> > --enable-linux-futex there), or just 4.4+
>
> I would rather put this in 4.3, especially considering the new work in
> this area from 4.2 hasn't been released yet. Then, we won't be changing
> it and changing it again.
Ok.
> As for the configure bits, I suggest using the same bits (exactly) for
> both libgomp and libstdc++, to avoid yet more twisty borrowing between
> these two runtimes.
>
> ;)
>
> Which means you might as well move it toplevel and avoid the
> duplication if possible. Thoughts?
Toplevel AFAIK doesn't contain configure tests like that, do you mean
just adding config/futex.m4 and using it in both libstdc++-v3 and libgomp
configury? That could theoretically work, there is just one obstackle
- libstdc++-v3 insists on doing no link tests when cross compiling
(don't understand why, I can understand why runtime tests aren't possible,
but link tests should be possible), while libgomp doesn't.
Jakub