From mboxrd@z Thu Jan  1 00:00:00 1970
From: Chris Wilson <chris@chris-wilson.co.uk>
Subject: Re: [PATCH 1/3] drm/i915: Wait for pending flips in
	intel_pipe_set_base()
Date: Fri, 02 Nov 2012 17:34:19 +0000
Message-ID: <b94cdc$76malg@fmsmga001.fm.intel.com>
References: <1351793163-8542-1-git-send-email-ville.syrjala@linux.intel.com>
	<1351793163-8542-2-git-send-email-ville.syrjala@linux.intel.com>
	<275ffc$785jr7@fmsmga002.fm.intel.com>
	<20121102172637.GV3791@intel.com>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="===============0125646362=="
Return-path: <intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org>
Received: from mga01.intel.com (mga01.intel.com [192.55.52.88])
	by gabe.freedesktop.org (Postfix) with ESMTP id 2D1E29E9FE
	for <intel-gfx@lists.freedesktop.org>;
	Fri,  2 Nov 2012 10:35:06 -0700 (PDT)
In-Reply-To: <20121102172637.GV3791@intel.com>
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
Cc: intel-gfx@lists.freedesktop.org
List-Id: intel-gfx@lists.freedesktop.org

--===============0125646362==
Content-Type: text/plain

On Fri, 2 Nov 2012 19:26:37 +0200, Ville Syrjälä <ville.syrjala@linux.intel.com> wrote:
> But now I started to wonder a bit about the performance implications
> of keeping struct_mutex locked for that long...

Unless you are worrying about doing a modeset for every vblank, don't.
We treat modesetting as a rare, stop-the-world event for the time being,
with the emphasis upon making the darn thing reliable. Long term goal
would be to decouple the locks so that we could have indepedent users
on a multi-head device, but for now we can assume that our sole user is
so distracted by the screen going blank to not worry about his
background folding@home task stalling for ~100 ms...

On the other hand, with output slaves it is likely that a render stall
will become more noticeable even for a single user.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre

--===============0125646362==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

--===============0125646362==--