From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ben Widawsky <ben@bwidawsk.net>
Subject: Re: [PATCH] [RFC] drm/i915: read-read semaphore
	optimization
Date: Tue, 13 Dec 2011 10:36:15 -0800
Message-ID: <4EE79B1F.2000707@bwidawsk.net>
References: <1323748328-10153-1-git-send-email-ben@bwidawsk.net>
	<87fwgoidza.fsf@eliezer.anholt.net>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; Format="flowed"
Content-Transfer-Encoding: 7bit
Return-path: <intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org>
Received: from cloud01.chad-versace.us (184-106-247-128.static.cloud-ips.com
	[184.106.247.128])
	by gabe.freedesktop.org (Postfix) with ESMTP id 1A51B9E755
	for <intel-gfx@lists.freedesktop.org>;
	Tue, 13 Dec 2011 10:36:28 -0800 (PST)
In-Reply-To: <87fwgoidza.fsf@eliezer.anholt.net>
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: Eric Anholt <eric@anholt.net>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>, intel-gfx@lists.freedesktop.org
List-Id: intel-gfx@lists.freedesktop.org

On 12/13/2011 09:22 AM, Eric Anholt wrote:
> On Mon, 12 Dec 2011 19:52:08 -0800, Ben Widawsky<ben@bwidawsk.net>  wrote:
>> Since we don't differentiate on the different GPU read domains, it
>> should be safe to allow back to back reads to occur without issuing a
>> wait (or flush in the non-semaphore case).
>>
>> This has the unfortunate side effect that we need to keep track of all
>> the outstanding buffer reads so that we can synchronize on a write, to
>> another ring (since we don't know which read finishes first). In other
>> words, the code is quite simple for two rings, but gets more tricky for
>>> 2 rings.
>>
>> Here is a picture of the solution to the above problem
>>
>> Ring 0            Ring 1             Ring 2
>> batch 0           batch 1            batch 2
>>    read buffer A     read buffer A      wait batch 0
>>                                         wait batch 1
>>                                         write buffer A
>>
>> This code is really untested. I'm hoping for some feedback if this is
>> worth cleaning up, and testing more thoroughly.
>
> You say it's an optimization -- do you have performance numbers?

33% improvement on a hacked version of gem_ring_sync_loop with.

It's not really a valid test as it's not coherent, but this is 
approximately the best case improvement.

Oddly semaphores doesn't make much difference in this test, which was 
surprising.