From mboxrd@z Thu Jan  1 00:00:00 1970
From: Chris Wilson <chris@chris-wilson.co.uk>
Subject: Re: [PATCH 11/18] drm/i915: switch to default context
	on idle
Date: Thu, 29 Mar 2012 21:28:10 +0100
Message-ID: <1333052905_146734@CP5-2952>
References: <1332103198-25852-1-git-send-email-ben@bwidawsk.net>
	<1332103198-25852-12-git-send-email-ben@bwidawsk.net>
	<20120329192905.GE27737@phenom.ffwll.local>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org>
Received: from fireflyinternet.com (smtp.fireflyinternet.com [109.228.6.236])
	by gabe.freedesktop.org (Postfix) with ESMTP id CA82B9E9DC
	for <intel-gfx@lists.freedesktop.org>;
	Thu, 29 Mar 2012 13:28:29 -0700 (PDT)
In-Reply-To: <20120329192905.GE27737@phenom.ffwll.local>
List-Unsubscribe: <http://lists.freedesktop.org/mailman/options/intel-gfx>,
	<mailto:intel-gfx-request@lists.freedesktop.org?subject=unsubscribe>
List-Archive: <http://lists.freedesktop.org/archives/intel-gfx>
List-Post: <mailto:intel-gfx@lists.freedesktop.org>
List-Help: <mailto:intel-gfx-request@lists.freedesktop.org?subject=help>
List-Subscribe: <http://lists.freedesktop.org/mailman/listinfo/intel-gfx>,
	<mailto:intel-gfx-request@lists.freedesktop.org?subject=subscribe>
Sender: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org
Errors-To: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org
To: Daniel Vetter <daniel@ffwll.ch>, Ben Widawsky <ben@bwidawsk.net>
Cc: intel-gfx@lists.freedesktop.org
List-Id: intel-gfx@lists.freedesktop.org

On Thu, 29 Mar 2012 21:29:05 +0200, Daniel Vetter <daniel@ffwll.ch> wrote:
> On Sun, Mar 18, 2012 at 01:39:51PM -0700, Ben Widawsky wrote:
> > To keep things as sane as possible, switch to the default context before
> > idling. This should help free context objects, as well as put things in
> > a more well defined state before suspending.
> > 
> > Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
> 
> Context are yet another thing that will nicely fragment our global gtt
> space. You don't pin them to mappable, which is a start, but I wonder
> whether we should integrate this into our evict_something code? Although
> that project is more involved with the current evict code, so maybe just
> add a fixme.

My suggestion for when we want to step beyond having the context pinned
all the time was for a hook into unbind that checked to see if this was
a "pinned" object and to fixup any registers to point back to the
default. In the case of contexts, it would check a flag to see if it was
indeed the active context and then schedule a change back to the default
context prior to unbinding. From another viewpoint, this sounds alot
like fences, and maybe we should add the tracking to the CXTID register
in a similar manner.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre