From mboxrd@z Thu Jan  1 00:00:00 1970
From: Julien Boulnois <jboulnois@gmail.com>
Subject: Re: omapdrm + etnaviv memory leak
Date: Wed, 23 May 2018 16:16:50 +0200
Message-ID: <CAB8EWQCJCTUJis-qt7+aMMfYTzD0fThG5i_mTGKE2O3RghXVSA@mail.gmail.com>
References: <ee1166e3-5423-b8f3-f7e5-dc7f87e156dc@ti.com>
 <1527065584.4227.10.camel@pengutronix.de>
 <6080b3f2-0fe6-dee7-331e-f2c0a9a3aa9d@ti.com>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="===============0573157618=="
Return-path: <dri-devel-bounces@lists.freedesktop.org>
Received: from mail-qk0-x22f.google.com (mail-qk0-x22f.google.com
 [IPv6:2607:f8b0:400d:c09::22f])
 by gabe.freedesktop.org (Postfix) with ESMTPS id 9F6CC6E4B6
 for <dri-devel@lists.freedesktop.org>; Wed, 23 May 2018 14:16:51 +0000 (UTC)
Received: by mail-qk0-x22f.google.com with SMTP id a8-v6so17544839qkj.12
 for <dri-devel@lists.freedesktop.org>; Wed, 23 May 2018 07:16:51 -0700 (PDT)
In-Reply-To: <6080b3f2-0fe6-dee7-331e-f2c0a9a3aa9d@ti.com>
List-Unsubscribe: <https://lists.freedesktop.org/mailman/options/dri-devel>,
 <mailto:dri-devel-request@lists.freedesktop.org?subject=unsubscribe>
List-Archive: <https://lists.freedesktop.org/archives/dri-devel>
List-Post: <mailto:dri-devel@lists.freedesktop.org>
List-Help: <mailto:dri-devel-request@lists.freedesktop.org?subject=help>
List-Subscribe: <https://lists.freedesktop.org/mailman/listinfo/dri-devel>,
 <mailto:dri-devel-request@lists.freedesktop.org?subject=subscribe>
Errors-To: dri-devel-bounces@lists.freedesktop.org
Sender: "dri-devel" <dri-devel-bounces@lists.freedesktop.org>
To: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: DRI Development <dri-devel@lists.freedesktop.org>
List-Id: dri-devel@lists.freedesktop.org

--===============0573157618==
Content-Type: multipart/alternative; boundary="0000000000009d9424056ce02f0a"

--0000000000009d9424056ce02f0a
Content-Type: text/plain; charset="UTF-8"

Hi Lucas,

2018-05-23 14:08 GMT+02:00 Tomi Valkeinen <tomi.valkeinen@ti.com>:

> On 23/05/18 11:53, Lucas Stach wrote:
> > Hi Tomi,
> >
> > Am Mittwoch, den 23.05.2018, 11:40 +0300 schrieb Tomi Valkeinen:
> >> Hi Lucas,
> >>
> >> Julien has written an X driver for OMAP5 SoC (which has Vivante's
> >> GC320). We're seeing a memory leak when using omapdrm for display and
> >> etnaviv for X 2D rendering. The X driver uses DRI3, so dmabuf
> >> import/export is involved.
> >
> > Why don't you extend the xf86-video-armada driver? It already supports
> > a load of Xrender and Compositing acceleration using the GC320. Adding
> > another scanout device shouldn't be hard.
>
> I don't know, I'll let Julien answer to that =). I think he had the
> armada driver working too.


 First of all, for fun and to educate myself ! Secondly, I found that
xf86-video-armada was pretty complex. I wanted to do something simpler
using existing higher level EXA API. Finally, when we started to work on
that with Tomi two years ago, the armada driver looked unmaintained and I
failed to get the DRI3/present part working. So it was actually easier to
work together using a brand new driver since we share the same hardware
(Beagleboard X15).

Regards

-- 
Julien Boulnois

--0000000000009d9424056ce02f0a
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi Lucas,<br><div><br>2018-05-23 14:08 GMT+02:00 Tomi Valk=
einen <span dir=3D"ltr">&lt;<a href=3D"mailto:tomi.valkeinen@ti.com" target=
=3D"_blank">tomi.valkeinen@ti.com</a>&gt;</span>:<br><div class=3D"gmail_ex=
tra"><div class=3D"gmail_quote"><blockquote class=3D"gmail_quote" style=3D"=
margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class=
=3D"">On 23/05/18 11:53, Lucas Stach wrote:<br>
&gt; Hi Tomi,<br>
&gt; <br>
&gt; Am Mittwoch, den 23.05.2018, 11:40 +0300 schrieb Tomi Valkeinen:<br>
&gt;&gt; Hi Lucas,<br>
&gt;&gt;<br>
&gt;&gt; Julien has written an X driver for OMAP5 SoC (which has Vivante&#3=
9;s<br>
&gt;&gt; GC320). We&#39;re seeing a memory leak when using omapdrm for disp=
lay and<br>
&gt;&gt; etnaviv for X 2D rendering. The X driver uses DRI3, so dmabuf<br>
&gt;&gt; import/export is involved.<br>
&gt; <br>
&gt; Why don&#39;t you extend the xf86-video-armada driver? It already supp=
orts<br>
&gt; a load of Xrender and Compositing acceleration using the GC320. Adding=
<br>
&gt; another scanout device shouldn&#39;t be hard.<br>
<br>
</span>I don&#39;t know, I&#39;ll let Julien answer to that =3D). I think h=
e had the<br>
armada driver working too.</blockquote><div><br></div><div>=C2=A0First of a=
ll, for fun and to educate myself ! Secondly, I found that<span class=3D"gm=
ail-"> xf86-video-armada</span> was pretty complex. I wanted to do somethin=
g simpler using existing higher level EXA API. Finally, when we started to =
work on that with Tomi two years ago, the armada driver looked unmaintained=
 and I failed to get the DRI3/present part working. So it was actually easi=
er to work together using a brand new driver since we share the same hardwa=
re (Beagleboard X15).<br><br></div><div>Regards<br></div></div><br>-- <br><=
div class=3D"gmail_signature" data-smartmail=3D"gmail_signature">Julien Bou=
lnois</div>
</div></div></div>

--0000000000009d9424056ce02f0a--

--===============0573157618==
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: base64
Content-Disposition: inline

X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KZHJpLWRldmVs
IG1haWxpbmcgbGlzdApkcmktZGV2ZWxAbGlzdHMuZnJlZWRlc2t0b3Aub3JnCmh0dHBzOi8vbGlz
dHMuZnJlZWRlc2t0b3Aub3JnL21haWxtYW4vbGlzdGluZm8vZHJpLWRldmVsCg==

--===============0573157618==--