[Buildroot] [PATCH RESEND v2] imx6q-sabresd: Add a qt5 defconfig variant

[Buildroot] [PATCH RESEND v2] imx6q-sabresd: Add a qt5 defconfig variant

[Fabio Estevam]

From: Fabio Estevam <festevam@gmail.com>

Introduce imx6q-sabresd_qt5_defconfig that supports the opensource
Etnaviv graphical stack.

This defconfig provides a way to quickly test some graphical applications,
such as kmscube, qt5, glmark2.

Currently kernel mainline exhibits issues when running cpufreq as ondemand
governor on mx6, so add a linux fragment that disables such option for the
time being.

Signed-off-by: Fabio Estevam <festevam@gmail.com>
---
Changes since v1:
- Use latest U-Boot and kernel versions

 board/freescale/imx6q-sabresd/linux.fragment       |  1 +
 board/freescale/imx6q-sabresd/readme.txt           | 23 +++++++++
 .../imx6q-sabresd/rootfs_overlay/root/sabresd.json | 15 ++++++
 configs/imx6q-sabresd_qt5_defconfig                | 59 ++++++++++++++++++++++
 4 files changed, 98 insertions(+)
 create mode 100644 board/freescale/imx6q-sabresd/linux.fragment
 create mode 100644 board/freescale/imx6q-sabresd/rootfs_overlay/root/sabresd.json
 create mode 100644 configs/imx6q-sabresd_qt5_defconfig

diff --git a/board/freescale/imx6q-sabresd/linux.fragment b/board/freescale/imx6q-sabresd/linux.fragment
new file mode 100644
index 0000000..39bec58
--- /dev/null
+++ b/board/freescale/imx6q-sabresd/linux.fragment
@@ -0,0 +1 @@
+# CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND is not set
diff --git a/board/freescale/imx6q-sabresd/readme.txt b/board/freescale/imx6q-sabresd/readme.txt
index 10e973c..b57c4fb 100644
--- a/board/freescale/imx6q-sabresd/readme.txt
+++ b/board/freescale/imx6q-sabresd/readme.txt
@@ -54,4 +54,27 @@ To boot your newly created system:
   emulator at 115200 bps, 8n1;
 - power on the board.
 
+Testing graphics on the i.MX6Q Sabre SD board
+=============================================
+
+The imx6q-sabresd_qt5_defconfig allows to quickly test the graphics
+capabilities of i.MX6 using the opensource Etnaviv graphics stack
+and kernel mainline.
+
+In order to build it:
+
+make imx6q-sabresd_qt5_defconfig
+make
+
+Then flash the SD card as explained above.
+
+Running kmscube application:
+
+# kmscube -D /dev/dri/card1
+
+Running Qt5 Cinematic Demo:
+
+# export QT_QPA_EGLFS_KMS_CONFIG=/root/sabresd.json
+# /usr/share/Qt5/CinematicExperience/Qt5_CinematicExperience
+
 Enjoy!
diff --git a/board/freescale/imx6q-sabresd/rootfs_overlay/root/sabresd.json b/board/freescale/imx6q-sabresd/rootfs_overlay/root/sabresd.json
new file mode 100644
index 0000000..9469dc1
--- /dev/null
+++ b/board/freescale/imx6q-sabresd/rootfs_overlay/root/sabresd.json
@@ -0,0 +1,15 @@
+{
+  "device": "/dev/dri/card1",
+  "hwcursor": false,
+  "pbuffers": true,
+  "outputs": [
+    {
+      "name": "HDMI-1",
+      "mode": "off"
+    },
+    {
+      "name": "LVDS-1",
+      "mode": "1024x768"
+    }
+  ]
+}
diff --git a/configs/imx6q-sabresd_qt5_defconfig b/configs/imx6q-sabresd_qt5_defconfig
new file mode 100644
index 0000000..f5e2686
--- /dev/null
+++ b/configs/imx6q-sabresd_qt5_defconfig
@@ -0,0 +1,59 @@
+BR2_arm=y
+BR2_cortex_a9=y
+BR2_ARM_ENABLE_NEON=y
+BR2_ARM_ENABLE_VFP=y
+BR2_ARM_FPU_VFPV3=y
+BR2_PACKAGE_HOST_LINUX_HEADERS_CUSTOM_4_11=y
+BR2_TOOLCHAIN_BUILDROOT_WCHAR=y
+BR2_TOOLCHAIN_BUILDROOT_CXX=y
+BR2_TOOLCHAIN_BUILDROOT_GLIBC=y
+BR2_TARGET_GENERIC_GETTY_PORT="ttymxc0"
+BR2_ROOTFS_POST_IMAGE_SCRIPT="board/freescale/common/imx/post-image.sh"
+BR2_LINUX_KERNEL=y
+BR2_LINUX_KERNEL_CUSTOM_VERSION=y
+BR2_LINUX_KERNEL_CUSTOM_VERSION_VALUE="4.11.4"
+BR2_LINUX_KERNEL_DEFCONFIG="imx_v6_v7"
+BR2_LINUX_KERNEL_DTS_SUPPORT=y
+BR2_LINUX_KERNEL_INTREE_DTS_NAME="imx6q-sabresd"
+BR2_LINUX_KERNEL_CONFIG_FRAGMENT_FILES="board/freescale/imx6q-sabresd/linux.fragment"
+BR2_ROOTFS_OVERLAY="board/freescale/imx6q-sabresd/rootfs_overlay"
+BR2_PACKAGE_STRACE=y
+BR2_PACKAGE_BITSTREAM_VERA=y
+BR2_PACKAGE_CANTARELL=y
+BR2_PACKAGE_DEJAVU=y
+BR2_PACKAGE_FONT_AWESOME=y
+BR2_PACKAGE_GHOSTSCRIPT_FONTS=y
+BR2_PACKAGE_INCONSOLATA=y
+BR2_PACKAGE_LIBERATION=y
+BR2_PACKAGE_GLMARK2=y
+BR2_PACKAGE_KMSCUBE=y
+BR2_PACKAGE_MESA3D_DEMOS=y
+BR2_PACKAGE_QT5CINEX=y
+BR2_PACKAGE_QT5CINEX_HD=y
+BR2_PACKAGE_MESA3D=y
+BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_ETNAVIV=y
+BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_SWRAST=y
+BR2_PACKAGE_MESA3D_DRI_DRIVER_SWRAST=y
+BR2_PACKAGE_MESA3D_OPENGL_ES=y
+BR2_PACKAGE_QT5=y
+BR2_PACKAGE_QT5BASE_LICENSE_APPROVED=y
+BR2_PACKAGE_QT5BASE_OPENGL_LIB=y
+BR2_PACKAGE_QT5BASE_LINUXFB=y
+BR2_PACKAGE_QT5BASE_FONTCONFIG=y
+BR2_PACKAGE_QT5BASE_GIF=y
+BR2_PACKAGE_QT5BASE_JPEG=y
+BR2_PACKAGE_FIRMWARE_IMX=y
+BR2_PACKAGE_INPUT_TOOLS=y
+BR2_PACKAGE_LIBV4L=y
+BR2_TARGET_ROOTFS_EXT2=y
+BR2_TARGET_ROOTFS_EXT2_4=y
+BR2_TARGET_ROOTFS_EXT2_BLOCKS=122880
+BR2_TARGET_UBOOT=y
+BR2_TARGET_UBOOT_BOARDNAME="mx6qsabresd"
+BR2_TARGET_UBOOT_CUSTOM_VERSION=y
+BR2_TARGET_UBOOT_CUSTOM_VERSION_VALUE="2017.05"
+BR2_TARGET_UBOOT_FORMAT_IMX=y
+BR2_PACKAGE_HOST_DOSFSTOOLS=y
+BR2_PACKAGE_HOST_GENIMAGE=y
+BR2_PACKAGE_HOST_MTOOLS=y
+BR2_PACKAGE_HOST_UBOOT_TOOLS=y
-- 
2.7.4

[Buildroot] [PATCH RESEND v2] imx6q-sabresd: Add a qt5 defconfig variant

[Thomas Petazzoni]

Hello,

On Wed, 21 Jun 2017 09:32:26 -0300, Fabio Estevam wrote:

> +BR2_PACKAGE_MESA3D=y
> +BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_ETNAVIV=y
> +BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_SWRAST=y
> +BR2_PACKAGE_MESA3D_DRI_DRIVER_SWRAST=y

Just curious, why are you enabling the swrast drivers here?

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

[Buildroot] [PATCH RESEND v2] imx6q-sabresd: Add a qt5 defconfig variant

[Fabio Estevam]

Hi Thomas,

On Wed, Jun 21, 2017 at 11:19 AM, Thomas Petazzoni
<thomas.petazzoni@free-electrons.com> wrote:
> Hello,
>
> On Wed, 21 Jun 2017 09:32:26 -0300, Fabio Estevam wrote:
>
>> +BR2_PACKAGE_MESA3D=y
>> +BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_ETNAVIV=y
>> +BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_SWRAST=y
>> +BR2_PACKAGE_MESA3D_DRI_DRIVER_SWRAST=y
>
> Just curious, why are you enabling the swrast drivers here?

Good point: actually I don't need it.

Would you like to respin it without this option?

Thanks

[Buildroot] [PATCH RESEND v2] imx6q-sabresd: Add a qt5 defconfig variant

[Thomas Petazzoni]

Hello,

On Wed, 21 Jun 2017 12:18:18 -0300, Fabio Estevam wrote:

> On Wed, Jun 21, 2017 at 11:19 AM, Thomas Petazzoni
> <thomas.petazzoni@free-electrons.com> wrote:
> > Hello,
> >
> > On Wed, 21 Jun 2017 09:32:26 -0300, Fabio Estevam wrote:
> >  
> >> +BR2_PACKAGE_MESA3D=y
> >> +BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_ETNAVIV=y
> >> +BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_SWRAST=y
> >> +BR2_PACKAGE_MESA3D_DRI_DRIVER_SWRAST=y  
> >
> > Just curious, why are you enabling the swrast drivers here?  
> 
> Good point: actually I don't need it.
> 
> Would you like to respin it without this option?

I can definitely remove them when applying, but I won't be able to do a
runtime test to verify that everything still works properly. So it
would probably be better if you could remove those lines, and verify
that things are still working as expected.

Thanks!

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

[Buildroot] [PATCH RESEND v2] imx6q-sabresd: Add a qt5 defconfig variant

[Fabio Estevam]

Hi Thomas,

On Wed, Jun 21, 2017 at 1:22 PM, Thomas Petazzoni
<thomas.petazzoni@free-electrons.com> wrote:

> I can definitely remove them when applying, but I won't be able to do a
> runtime test to verify that everything still works properly. So it
> would probably be better if you could remove those lines, and verify
> that things are still working as expected.

I have tested and it works fine. Just sent a v3, thanks.

[Buildroot] [PATCH RESEND v2] imx6q-sabresd: Add a qt5 defconfig variant

[Fabio Estevam]

Hi Thomas,

On Wed, Jun 21, 2017 at 5:50 PM, Fabio Estevam <festevam@gmail.com> wrote:
> Hi Thomas,
>
> On Wed, Jun 21, 2017 at 1:22 PM, Thomas Petazzoni
> <thomas.petazzoni@free-electrons.com> wrote:
>
>> I can definitely remove them when applying, but I won't be able to do a
>> runtime test to verify that everything still works properly. So it
>> would probably be better if you could remove those lines, and verify
>> that things are still working as expected.
>
> I have tested and it works fine. Just sent a v3, thanks.

Are you happy with v3?