From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from gateway24.websitewelcome.com (gateway24.websitewelcome.com [192.185.51.196]) by mx.groups.io with SMTP id smtpd.web12.527.1576017146697446038 for ; Tue, 10 Dec 2019 14:32:27 -0800 Authentication-Results: mx.groups.io; dkim=fail reason="no key for verify" header.i=@montgomery1.com header.s=default header.b=SI7w2D9i; spf=permerror, err=parse error for token &{10 18 bluehost.com}: parse error for token &{10 18 _spf.google.com}: parse error for token &{10 18 _netblocks3.google.com}: limit exceeded (domain: montgomery1.com, ip: 192.185.51.196, mailfrom: clay@montgomery1.com) Received: from cm17.websitewelcome.com (cm17.websitewelcome.com [100.42.49.20]) by gateway24.websitewelcome.com (Postfix) with ESMTP id 1A9051F86 for ; Tue, 10 Dec 2019 16:32:26 -0600 (CST) Received: from box5869.bluehost.com ([162.241.24.119]) by cmsmtp with SMTP id eo3Nil1bdqNtveo3NiU12H; Tue, 10 Dec 2019 16:32:26 -0600 X-Authority-Reason: nr=8 DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=montgomery1.com; s=default; h=Content-Type:In-Reply-To:MIME-Version:Date: Message-ID:From:References:To:Subject:Sender:Reply-To:Cc: Content-Transfer-Encoding:Content-ID:Content-Description:Resent-Date: Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Id: List-Help:List-Unsubscribe:List-Subscribe:List-Post:List-Owner:List-Archive; bh=PD5dRNw96tw8NvXLuP3PJRmpKTbJRwhsEUcVRiqK7d0=; b=SI7w2D9isw5Gb0HGgJGw+h9iX iXnmlCFQXQYowKWebu/8K+1cNUKEv915huUWvTJy6Clwc5rKKhZdYs+LX3di81vV+58gUY5Z51Bmt 4VcVi6zCTH9EooabY47Dei9yjt8f4WpTvg5eaLRjX1UmRim0qRqrsss6Me6v1A/rBYvDYIqZ/cIGs Tu8OG17G/GPOaSWlWB+NJNPmk+9VbklaAx1Cimou/3XD9Qjpptd146t6aNl6RGj7bpJTdEZriT7qp LXhljx2vvvxAXH0XbI/hQWLd163s8s8vlA6YX8gJ9+8NCDpc76XQsXTRWCxUGxnhvFkx5pJn0Dro1 NvtkdOvvw==; Received: from [97.99.192.131] (port=55722 helo=[192.168.0.4]) by box5869.bluehost.com with esmtpsa (TLSv1.2:ECDHE-RSA-AES128-GCM-SHA256:128) (Exim 4.92) (envelope-from ) id 1ieo3N-002m7s-Mi; Tue, 10 Dec 2019 15:32:25 -0700 Subject: Re: [meta-freescale] How to get an fb device with Sumo on i.MX6? To: Martin Fuzzey , meta-freescale@lists.yoctoproject.org References: <3fada3b8-5b00-ef8c-8849-b2a04031b4dc@montgomery1.com> <1dab305c-3cc4-9b62-d374-44c5fdea0579@montgomery1.com> From: "Clay Montgomery" Organization: Montgomery One Message-ID: Date: Tue, 10 Dec 2019 16:32:30 -0600 User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.3.0 MIME-Version: 1.0 In-Reply-To: X-AntiAbuse: This header was added to track abuse, please include it with any abuse report X-AntiAbuse: Primary Hostname - box5869.bluehost.com X-AntiAbuse: Original Domain - lists.yoctoproject.org X-AntiAbuse: Originator/Caller UID/GID - [47 12] / [47 12] X-AntiAbuse: Sender Address Domain - montgomery1.com X-BWhitelist: no X-Source-IP: 97.99.192.131 X-Source-L: No X-Exim-ID: 1ieo3N-002m7s-Mi X-Source: X-Source-Args: X-Source-Dir: X-Source-Sender: ([192.168.0.4]) [97.99.192.131]:55722 X-Source-Auth: clay@montgomery1.com X-Email-Count: 2 X-Source-Cap: bW9udGdvcjA7bW9udGdvcjA7Ym94NTg2OS5ibHVlaG9zdC5jb20= X-Local-Domain: yes Content-Type: multipart/alternative; boundary="------------8A818141B47EDB4148B287FC" Content-Language: en-US --------------8A818141B47EDB4148B287FC Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: quoted-printable Martin, Otavio, Andrey, I did some more testing and found the situation is actually worse than I=20 thought. Even the X11 and Wayland support is completely non-functional,=20 in addition to fb, on the i.MX6. The linux kernel version makes no=20 difference. I have tried 4.8 to 4.14. It's obvious that NXP is not even=20 testing on the i.MX6 anymore, despite what their documentation and=20 release notes claim. =C2=A0This means the Vivante 6.2.4 package is COMPLETELY USELESS on the=20 i.MX6. If there is anyone who has actually gotten this to work in any=20 mode, please reply. =C2=A0I will look at Etnaviv again, but I doubt it provides the OpenGL E= S=20 extensions required to accelerate video, or OpenCL, or the VDK with=20 emulator, performance tuning tools, etc. If NXP never provides another release that is functional on i.MX6, then=20 mainline and Etnaviv is the only way to use kernels past 4.1.15. Thank you for your assistance. Regards, Clay On 12/4/2019 3:33 PM, Martin Fuzzey wrote: > I would strongly advise using any NXP code. > Use latest mainline kernel DRM drivers and etnaviv if you need GPU. > > The NXP framebuffer has been a dead end for years as it doesn't=20 > integrate into the DRM subsystem. > > Note if you need a classic framebuffer interface to userspace for some=20 > reason (/dev/fb) you still get that with fbdev emulation on top of DRM. > > On Tue, 3 Dec 2019, 22:00 Clay Montgomery, > wrote: > > Hi Andrey, > > =C2=A0=C2=A0=C2=A0 I would greatly appreciate any assistance you ca= n provide. > > I just tested two more kernel versions, 4.9.67 and 4.8.17 and > found that > the fb device is broken on these also. > This means this issue is older than I thought. As best as I can tel= l, > NXP has not released a BSP that actually works for > DISTRO=3D'fsl-imx-fb' > on the i.MX6 for almost 3 years. The last release that I know > works was > Pyro, but that is using the 5.0.11.p8 version of the Vivante packag= e, > which was released in May, 2017. > > This issue may be complicated by incorrect settings of the kernel > config > options. I have listed what I am using below, which worked with Pyr= o. > > Thanks, Clay > > > > On 12/3/2019 2:57 AM, Andrey Zhizhikin wrote: > > Hello Clay, > > > > On Tue, Dec 3, 2019 at 2:31 AM Clay Montgomery > > wrote: > >> > >> I will share what I found about the broken fb device issue on > the i.MX6. > >> > >> I got a reply from a developer within NXP that indicates there i= s a > >> known bug with their fb device support with the 4.14.98 kernel. > It seems > >> that the last kernel where the fb works is probably 4.9.51, > which they > >> recommend. > > Do you have any additional information here to share about the > kernel > > issue? I was looking into updating the 4.9.y branch of > linux-freescale > > and could look into this as well. Perhaps there are commits from = NXP > > that I can isolate to have this resolved... > > > >> I suspect this may be the same issue which caused NXP to decide > to drop > >> all support for fb (and X11) on the i.MX8. However, this issue > affects > >> the i.MX6 also and I see no indication that it will ever be > fixed, since > >> NXP's plan seems to be to support Wayland only, at least on the > i.MX8. > >> > >> This seems strange to me because it freezes development on the > i.MX6 at > >> the 4.9.51 kernel, unless Wayland is used. Even a lot of NXP's > own unit > >> tests for the GPU, VPU, IPU and V4L2 will not run without fb, ev= en > >> though their release notes for their Yocto support claim they > tested on > >> a lot of boards. It seems that testing merely means booting and = not > >> actually running unit tests. > >> > >> If I have this wrong or anyone can shed more light on this > situation, > >> please do. > >> > >> Thanks, Clay Montgomery > >> > >> > >> > >> On 11/24/2019 2:54 PM, Clay D. Montgomery wrote: > >>> Hello, > >>> > >>>=C2=A0 =C2=A0 =C2=A0I am trying to migrate a project from Yocto = Pyro (3.14 > kernel) to > >>> Sumo and a newer kernel on the i.MX6 Wandboard Quad. > >>> > >>> I am following the instructions in the > >>> i.MX_Yocto_Project_User's_Guide, to build. To summarize, I am > using: > >>> > >>>=C2=A0 =C2=A0 =C2=A0 repo init=C2=A0 -u > >>> https://source.codeaurora.org/external/imx/imx-manifest -b > >>> imx-linux-sumo=C2=A0 -m=C2=A0 imx-4.14.98-2.0.0_ga.xml > >>>=C2=A0 =C2=A0 =C2=A0 DISTRO=3Dfsl-imx-fb=C2=A0 MACHINE=3Dwandboa= rd=C2=A0 source > fsl-setup-release.sh > >>> -b=C2=A0 build-fb > >>>=C2=A0 =C2=A0 =C2=A0 bitbake=C2=A0 fsl-image-machine-test > >>> > >>> The problem is that the built linux image has no fb device and > this > >>> causes many other problems, such as with V4L2, OpenGL ES and ev= en > >>> NXP's unit tests, such as: > >>> > >>>=C2=A0 =C2=A0 =C2=A0 ./mxc_vpu_test -D "-f0 -x0 -i akiyo.mp4" > >>> > >>>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 [ERR]=C2=A0 =C2=A0unable to o= pen fb0 > >>> > >>> Even 'fbset' fails, with: > >>>=C2=A0 =C2=A0 =C2=A0 open /dev/fb0: No such file or directory > >>> > >>> As best as I can tell, fb support should be enabled. Here is th= e > >>> relevant portion of my defconfig: > >>> > >>> CONFIG_DRM=3Dy > >>> CONFIG_DRM_VIVANTE=3Dy > >>> CONFIG_FB=3Dy > >>> CONFIG_FB_MXS=3Dy > >>> CONFIG_FB_MXC_SYNC_PANEL=3Dy > >>> CONFIG_FB_MXC_MIPI_DSI=3Dy > >>> CONFIG_FB_MXC_MIPI_DSI_SAMSUNG=3Dy > >>> CONFIG_FB_MXC_TRULY_WVGA_SYNC_PANEL=3Dy > >>> CONFIG_FB_MXC_TRULY_PANEL_TFT3P5079E=3Dy > >>> CONFIG_FB_MXC_TRULY_PANEL_TFT3P5581E=3Dy > >>> CONFIG_FB_MXC_LDB=3Dy > >>> CONFIG_FB_MXC_HDMI=3Dy > >>> CONFIG_FB_MXS_SII902X=3Dy > >>> CONFIG_FB_MXC_DCIC=3Dm > >>> CONFIG_HANNSTAR_CABC=3Dy > >>> CONFIG_FB_MXC_EINK_PANEL=3Dy > >>> CONFIG_FB_MXC_EINK_V2_PANEL=3Dy > >>> CONFIG_LCD_CLASS_DEVICE=3Dy > >>> CONFIG_LCD_L4F00242T03=3Dy > >>> CONFIG_LCD_PLATFORM=3Dy > >>> CONFIG_BACKLIGHT_PWM=3Dy > >>> CONFIG_FRAMEBUFFER_CONSOLE=3Dy > >>> CONFIG_FRAMEBUFFER_CONSOLE_DETECT_PRIMARY=3Dy > >>> CONFIG_MXC_IPU=3Dy > >>> CONFIG_MXC_IPU_V3_PRE=3Dy > >>> CONFIG_MXC_SIM=3Dy > >>> CONFIG_MXC_MIPI_CSI2=3Dy > >>> CONFIG_MXC_HDMI_CEC=3Dy > >>> CONFIG_MXC_GPU_VIV=3Dy > >>> > >>> I have found that the fb device will work if I select the > >>> 'CONFIG_DRM_ETNAVIV' option, instead of 'CONFIG_DRM_VIVANTE', > but I > >>> need the Vivante option to get their newer drivers. > >>> > >>>=C2=A0 =C2=A0 Can someone please advise how I can resolve this i= ssue? > >>> > >>>=C2=A0 =C2=A0 Any assistance would be greatly appreciated. > >>> > >>> Thank You, Clay Montgomery > >>> > >>> > >> -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D- > >> Links: You receive all messages sent to this group. > >> > >> View/Reply Online (#24145): > https://lists.yoctoproject.org/g/meta-freescale/message/24145 > >> Mute This Topic: https://lists.yoctoproject.org/mt/61881553/3617= 192 > >> Group Owner: meta-freescale+owner@lists.yoctoproject.org > > >> Unsubscribe: > https://lists.yoctoproject.org/g/meta-freescale/unsub > [andrey.z@gmail.com ] > >> -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D- > > > > > -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D- > Links: You receive all messages sent to this group. > > View/Reply Online (#24147): > https://lists.yoctoproject.org/g/meta-freescale/message/24147 > Mute This Topic: https://lists.yoctoproject.org/mt/61881553/3617492 > Group Owner: meta-freescale+owner@lists.yoctoproject.org > > Unsubscribe: https://lists.yoctoproject.org/g/meta-freescale/unsub > [mfuzzey@gmail.com ] > -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D- > > > -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D- > Links: You receive all messages sent to this group. > > View/Reply Online (#24149): https://lists.yoctoproject.org/g/meta-frees= cale/message/24149 > Mute This Topic: https://lists.yoctoproject.org/mt/61881553/3617246 > Group Owner: meta-freescale+owner@lists.yoctoproject.org > Unsubscribe: https://lists.yoctoproject.org/g/meta-freescale/unsub [cl= ay@montgomery1.com] > -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D- --------------8A818141B47EDB4148B287FC Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable

Martin, Otavio, Andrey,

I did some more testing and found the situation is actually worse than I thought. Even the X11 and Wayland support is completely non-functional, in addition to fb, on the i.MX6. The linux kernel version makes no difference. I have tried 4.8 to 4.14. It's obvious that NXP is not even testing on the i.MX6 anymore, despite what their documentation and release notes claim.

=C2=A0This means the Vivante 6.2.4 package is COMPLETELY USELESS o= n the i.MX6. If there is anyone who has actually gotten this to work in any mode, please reply.

=C2=A0I will look at Etna= viv again, but I doubt it provides the OpenGL ES extensions required to accelerate video, or OpenCL, or the VDK with emulator, performance tuning tools, etc.

If NXP never provides another release that is functional on i.MX6, then mainline and Etnaviv is the only way to use kernels past 4.1.15.

Thank you for your assistance.

Regards, Clay


On 12/4/2019 3:33 PM, Martin Fuzzey wrote:
I would strongly advise using any NXP code.
Use latest mainline kernel DRM drivers and etnaviv if you need GPU.

The NXP framebuffer has been a dead end for years as it doesn't integrate into the DRM subsystem.

Note if you need a classic framebuffer interfac= e to userspace for some reason (/dev/fb) you still get that with fbdev emulation on top of DRM.=C2=A0

On Tue, 3 Dec 2019, 22:00 C= lay Montgomery, <clay@montgomery1.com> wrote:<= br>
Hi Andrey,
=C2=A0=C2=A0=C2=A0 I would greatly appreciate any assistance yo= u can provide.

I just tested two more kernel versions, 4.9.67 and 4.8.17 and found that
the fb device is broken on these also.
This means this issue is older than I thought. As best as I can tell,
NXP has not released a BSP that actually works for DISTRO=3D'fsl-imx-fb'
on the i.MX6 for almost 3 years. The last release that I know works was
Pyro, but that is using the 5.0.11.p8 version of the Vivante package,
which was released in May, 2017.

This issue may be complicated by incorrect settings of the kernel config
options. I have listed what I am using below, which worked with Pyro.

Thanks, Clay



On 12/3/2019 2:57 AM, Andrey Zhizhikin wrote:
> Hello Clay,
>
> On Tue, Dec 3, 2019 at 2:31 AM Clay Montgomery <clay@montgomery1.= com> wrote:
>>
>> I will share what I found about the broken fb device issue on the i.MX6.
>>
>> I got a reply from a developer within NXP that indicates there is a
>> known bug with their fb device support with the 4.14.98 kernel. It seems
>> that the last kernel where the fb works is probably 4.9.51, which they
>> recommend.
> Do you have any additional information here to share about the kernel
> issue? I was looking into updating the 4.9.y branch of linux-freescale
> and could look into this as well. Perhaps there are commits from NXP
> that I can isolate to have this resolved...
>
>> I suspect this may be the same issue which caused NXP to decide to drop
>> all support for fb (and X11) on the i.MX8. However, this issue affects
>> the i.MX6 also and I see no indication that it will ever be fixed, since
>> NXP's plan seems to be to support Wayland only, at least on the i.MX8.
>>
>> This seems strange to me because it freezes development on the i.MX6 at
>> the 4.9.51 kernel, unless Wayland is used. Even a lot of NXP's own unit
>> tests for the GPU, VPU, IPU and V4L2 will not run without fb, even
>> though their release notes for their Yocto support claim they tested on
>> a lot of boards. It seems that testing merely means booting and not
>> actually running unit tests.
>>
>> If I have this wrong or anyone can shed more light on this situation,
>> please do.
>>
>> Thanks, Clay Montgomery
>>
>>
>>
>> On 11/24/2019 2:54 PM, Clay D. Montgomery wrote:
>>> Hello,
>>>
>>>=C2=A0 =C2=A0 =C2=A0I am trying to migrate a projec= t from Yocto Pyro (3.14 kernel) to
>>> Sumo and a newer kernel on the i.MX6 Wandboard Quad.
>>>
>>> I am following the instructions in the
>>> i.MX_Yocto_Project_User's_Guide, to build. To summarize, I am using:
>>>
>>>=C2=A0 =C2=A0 =C2=A0 repo init=C2=A0 -u
>>> https://source.codeaurora.org/extern= al/imx/imx-manifest=C2=A0 -b
>>> imx-linux-sumo=C2=A0 -m=C2=A0 imx-4.14.98-2.0.0_ga= .xml
>>>=C2=A0 =C2=A0 =C2=A0 DISTRO=3Dfsl-imx-fb=C2=A0 MACH= INE=3Dwandboard=C2=A0 source fsl-setup-release.sh
>>> -b=C2=A0 build-fb
>>>=C2=A0 =C2=A0 =C2=A0 bitbake=C2=A0 fsl-image-machin= e-test
>>>
>>> The problem is that the built linux image has no fb device and this
>>> causes many other problems, such as with V4L2, OpenGL ES and even
>>> NXP's unit tests, such as:
>>>
>>>=C2=A0 =C2=A0 =C2=A0 ./mxc_vpu_test -D "-f0 -x0 -i = akiyo.mp4"
>>>
>>>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 [ERR]=C2=A0 =C2=A0= unable to open fb0
>>>
>>> Even 'fbset' fails, with:
>>>=C2=A0 =C2=A0 =C2=A0 open /dev/fb0: No such file or= directory
>>>
>>> As best as I can tell, fb support should be enabled. Here is the
>>> relevant portion of my defconfig:
>>>
>>> CONFIG_DRM=3Dy
>>> CONFIG_DRM_VIVANTE=3Dy
>>> CONFIG_FB=3Dy
>>> CONFIG_FB_MXS=3Dy
>>> CONFIG_FB_MXC_SYNC_PANEL=3Dy
>>> CONFIG_FB_MXC_MIPI_DSI=3Dy
>>> CONFIG_FB_MXC_MIPI_DSI_SAMSUNG=3Dy
>>> CONFIG_FB_MXC_TRULY_WVGA_SYNC_PANEL=3Dy
>>> CONFIG_FB_MXC_TRULY_PANEL_TFT3P5079E=3Dy
>>> CONFIG_FB_MXC_TRULY_PANEL_TFT3P5581E=3Dy
>>> CONFIG_FB_MXC_LDB=3Dy
>>> CONFIG_FB_MXC_HDMI=3Dy
>>> CONFIG_FB_MXS_SII902X=3Dy
>>> CONFIG_FB_MXC_DCIC=3Dm
>>> CONFIG_HANNSTAR_CABC=3Dy
>>> CONFIG_FB_MXC_EINK_PANEL=3Dy
>>> CONFIG_FB_MXC_EINK_V2_PANEL=3Dy
>>> CONFIG_LCD_CLASS_DEVICE=3Dy
>>> CONFIG_LCD_L4F00242T03=3Dy
>>> CONFIG_LCD_PLATFORM=3Dy
>>> CONFIG_BACKLIGHT_PWM=3Dy
>>> CONFIG_FRAMEBUFFER_CONSOLE=3Dy
>>> CONFIG_FRAMEBUFFER_CONSOLE_DETECT_PRIMARY=3Dy
>>> CONFIG_MXC_IPU=3Dy
>>> CONFIG_MXC_IPU_V3_PRE=3Dy
>>> CONFIG_MXC_SIM=3Dy
>>> CONFIG_MXC_MIPI_CSI2=3Dy
>>> CONFIG_MXC_HDMI_CEC=3Dy
>>> CONFIG_MXC_GPU_VIV=3Dy
>>>
>>> I have found that the fb device will work if I select the
>>> 'CONFIG_DRM_ETNAVIV' option, instead of 'CONFIG_DRM_VIVANTE', but I
>>> need the Vivante option to get their newer drivers.
>>>
>>>=C2=A0 =C2=A0 Can someone please advise how I can r= esolve this issue?
>>>
>>>=C2=A0 =C2=A0 Any assistance would be greatly appre= ciated.
>>>
>>> Thank You, Clay Montgomery
>>>
>>>
>> -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-
>> Links: You receive all messages sent to this group. >>
>> View/Reply Online (#24145): https://lists.yoctoproject.org/g/met= a-freescale/message/24145
>> Mute This Topic: https://lists.yoctoproject.org/mt/61= 881553/3617192
>> Group Owner: meta-freescale+owner@lists.yoctoproject.org
>> Unsubscribe: https://lists.yoctoproject.org/g/met= a-freescale/unsub=C2=A0 [andrey.z@gmail.co= m]
>> -=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-
>
>
-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-
Links: You receive all messages sent to this group.

View/Reply Online (#24147): https://lists.yoctoproject.org/g/met= a-freescale/message/24147
Mute This Topic: https://lists.yoctoproject.org/mt/61= 881553/3617492
Group Owner: meta-freescale+owner@lists.yoctoproject.org
Unsubscribe: https://lists.yoctoproject.org/g/met= a-freescale/unsub=C2=A0 [mfuzzey@gmail.com= ]
-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-

-=3D-=3D-=3D-=3D-=3D-=3D-=3D=
-=3D-=3D-=3D-=3D-
Links: You receive all messages sent to this group.

View/Reply Online (#24149): https://list=
s.yoctoproject.org/g/meta-freescale/message/24149
Mute This Topic: https://lists.yoctoproject.org/mt/=
61881553/3617246
Group Owner: meta-freescale+owner@lists.yoctopro=
ject.org
Unsubscribe: https://lists.yoctoproject.org/g/me=
ta-freescale/unsub  [clay@montgomery1.com]
-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-
--------------8A818141B47EDB4148B287FC--