From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jerome Glisse <j.glisse@gmail.com>
Subject: Re: [RFC] Heterogeneous memory management (mirror process address
 space on a device mmu).
Date: Tue, 6 May 2014 11:00:17 -0400
Message-ID: <20140506150014.GA6731@gmail.com>
References: <1399038730-25641-1-git-send-email-j.glisse@gmail.com>
 <20140506102925.GD11096@twins.programming.kicks-ass.net>
 <CA+55aFzt47Jpp-KK-ocLGgzYt_w-vheqFLfaGZOUSjwVrgGUtw@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable
Cc: Peter Zijlstra <peterz@infradead.org>, linux-mm <linux-mm@kvack.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>,
	Mel Gorman <mgorman@suse.de>, "H. Peter Anvin" <hpa@zytor.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Linda Wang <lwang@redhat.com>, Kevin E Martin <kem@redhat.com>,
	Jerome Glisse <jglisse@redhat.com>,
	Andrea Arcangeli <aarcange@redhat.com>,
	Johannes Weiner <jweiner@redhat.com>,
	Larry Woodman <lwoodman@redhat.com>, Rik van Riel <riel@redhat.com>,
	Dave Airlie <airlied@redhat.com>, Jeff Law <law@redhat.com>,
	Brendan Conoboy <blc@redhat.com>, Joe Donohue <jdonohue@redhat.com>,
	Duncan Poole <dpoole@nvidia.com>,
	Sherry Cheung <SCheung@nvidia.com>,
	Subhash Gutti <sgutti@nvidia.com>,
	John Hubbard <jhubbard@nvidia.com>,
	Mark Hairgrove <mhairgrove@nvidia.com>,
	Lucien Dunning <ldunning@nvidia.com>,
To: Linus Torvalds <torvalds@linux-foundation.org>
Return-path: <owner-linux-mm@kvack.org>
Content-Disposition: inline
In-Reply-To: <CA+55aFzt47Jpp-KK-ocLGgzYt_w-vheqFLfaGZOUSjwVrgGUtw@mail.gmail.com>
Sender: owner-linux-mm@kvack.org
List-Id: linux-fsdevel.vger.kernel.org

On Tue, May 06, 2014 at 07:57:02AM -0700, Linus Torvalds wrote:
> On Tue, May 6, 2014 at 3:29 AM, Peter Zijlstra <peterz@infradead.org> w=
rote:
> >
> > So you forgot to CC Linus, Linus has expressed some dislike for
> > preemptible mmu_notifiers in the recent past:
>=20
> Indeed. I think we *really* should change that anonvma rwsem into an
> rwlock. We had performance numbers that showed it needs to be done.
>=20
> The *last* thing we want is to have random callbacks that can block in
> this critical region. So now I think making it an rwlock is a good
> idea just to make sure that never happens.
>=20
> Seriously, the mmu_notifiers were misdesigned to begin with, and much
> too deep. We're not screwing up the VM any more because of them.
>=20
>                  Linus

So question becomes how to implement process address space mirroring
without pinning memory and track cpu page table update knowing that
device page table update is unbound can not be atomic from cpu point
of view.

Cheers,
J=E9r=F4me

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=3Dmailto:"dont@kvack.org"> email@kvack.org </a>