Device Tree display-subsystem and Frame Buffer Creation

[asira]

Hi everyone,

I had some technical questions about how the display system gets defined using the device tree. I have a parallel interface LCD available through /dev/fb0. I'm trying to figure out where fb0 comes from in the device tree so I can adjust timings and so on with a display-timings block.

So, in imx6q.dtsi there's this code:

imx6q.dtsi

ipu2_di1: port@3 {             #address-cells = <1>;             #size-cells = <0>;             reg = <3>;             ipu2_di1_hdmi: endpoint@1 {                 remote-endpoint = <&hdmi_mux_3>;             };             ipu2_di1_mipi: endpoint@2 {             };             ipu2_di1_lvds0: endpoint@3 {                 remote-endpoint = <&lvds0_mux_3>;             };             ipu2_di1_lvds1: endpoint@4 {                 remote-endpoint = <&lvds1_mux_3>;             };         };     }; }; display-subsystem {     compatible = "fsl,imx-display-subsystem";     ports = <&ipu1_di0>, <&ipu1_di1>, <&ipu2_di0>, <&ipu2_di1>; };

ipu1 is defined in imx6qdl.dtsi. The display-subsystem section is of particular interest to me. When I run dmesg get this:

debian@arm:~$ sudo dmesg | grep fb0
[ 3.888801] imx-drm display-subsystem: fb0: frame buffer device

The complete set of messages is this:

[ 3.758204] imx-ipuv3 2400000.ipu: IPUv3H probed
[ 3.764282] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
[ 3.770951] [drm] No driver support for vblank timestamp query.
[ 3.777006] imx-drm display-subsystem: bound imx-ipuv3-crtc.0 (ops ipu_crtc_ops)
[ 3.784480] imx-drm display-subsystem: bound imx-ipuv3-crtc.1 (ops ipu_crtc_ops)
[ 3.791986] imx-drm display-subsystem: bound imx-ipuv3-crtc.4 (ops ipu_crtc_ops)
[ 3.799492] imx-drm display-subsystem: bound imx-ipuv3-crtc.5 (ops ipu_crtc_ops)
[ 3.807121] dwhdmi-imx 120000.hdmi: Detected HDMI controller 0x13:0xa:0xa0:0xc1
[ 3.814641] [drm] forcing HDMI-A-1 connector ON
[ 3.819228] imx-drm display-subsystem: bound 120000.hdmi (ops dw_hdmi_imx_ops)
[ 3.877144] Console: switching to colour frame buffer device 128x48
[ 3.888801] imx-drm display-subsystem: fb0: frame buffer device
[ 3.894835] imx-drm display-subsystem: registered panic notifier
[ 3.954779] [drm] Initialized imx-drm 1.0.0 20120507 on minor 0
[ 3.961116] imx-ipuv3 2800000.ipu: IPUv3H probed
 

So, I'm wondering where does /dev/fb0 come from in the device tree using the default configuration? There's no fb statement like what's shown here as an example. I need to adjust the clock polarity and other timing stuff for the display. How can I go about doing that properly?

Thanks!

[igorpadykov]

Hi Ashwin

seems in default configuration fb number is configured during

registering in mxcfb_probe() function, please check
mxc_dispdrv.c,mxc_lcdif.c,mxc_ipuv3_fb.c and Table 6-1. IPU Driver Files

in attached Linux Manual.

Best regards
igor
-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

[asira]

Looks helpful. I'll read through it. I discovered that the location of the status="okay"; statement matters, and now with that statement under the configuration for that node I get the following out of dmesg:

[ 3.776730] imx-drm display-subsystem: bound imx-ipuv3-crtc.0 (ops ipu_crtc_ops)
[ 3.784208] imx-drm display-subsystem: bound imx-ipuv3-crtc.1 (ops ipu_crtc_ops)
[ 3.791721] imx-drm display-subsystem: bound imx-ipuv3-crtc.4 (ops ipu_crtc_ops)
[ 3.799226] imx-drm display-subsystem: bound imx-ipuv3-crtc.5 (ops ipu_crtc_ops)
[ 3.806843] dwhdmi-imx 120000.hdmi: Detected HDMI controller 0x13:0xa:0xa0:0xc1
[ 3.814351] [drm] forcing HDMI-A-1 connector ON
[ 3.818937] imx-drm display-subsystem: bound 120000.hdmi (ops dw_hdmi_imx_ops)
[ 3.826355] imx-ldb 2000000.aips-bus:ldb@020e0008: could not determine data mapping: -22
[ 3.834470] imx-drm display-subsystem: failed to bind 2000000.aips-bus:ldb@020e0008 (ops imx_ldb_ops): -22
[ 3.844925] imx-drm display-subsystem: master bind failed: -22
[ 3.850787] imx-ipuv3-crtc: probe of imx-ipuv3-crtc.5 failed with error -22
[ 3.858174] imx-ipuv3 2800000.ipu: IPUv3H probed

Which is an improvement I guess, it's at least trying to do something. I haven't found a translation for error -22 yet though. The LDB declaration I'm making on top of the default code is:

&ldb {

    interface_pix_fmt = "IPU_PIX_FMT_RGB666";

    clocks = <&clks IMX6QDL_CLK_LDB_DI0_SEL>, <&clks IMX6QDL_CLK_LDB_DI1_SEL>,
        <&clks IMX6QDL_CLK_IPU1_DI0_SEL>, <&clks IMX6QDL_CLK_IPU1_DI1_SEL>,
       <&clks IMX6QDL_CLK_IPU2_DI0_SEL>, <&clks IMX6QDL_CLK_IPU2_DI1_SEL>,
       <&clks IMX6QDL_CLK_LDB_DI0>, <&clks IMX6QDL_CLK_LDB_DI1>,
       <&clks IMX6QDL_CLK_LDB_DI0_DIV_3_5>, <&clks IMX6QDL_CLK_LDB_DI1_DIV_3_5>,
       <&clks IMX6QDL_CLK_LDB_DI0_SEL>, <&clks IMX6QDL_CLK_LDB_DI1_SEL>;

    clock-names = "di0_pll", "di1_pll",
                  "di0_sel", "di1_sel",
                  "di2_sel", "di3_sel",
                  "di0", "di1",
                  "ldb_di0_div_3_5", "ldb_di1_div_3_5",
                  "ldb_di0_sel","ldb_di1_sel";

    status = "okay";

    lvds-channel@0 {

        crtc = "ipu1-di1";
        interface_pix_fmt = "RGB666";
        status = "okay";

        display-timings {
            native-mode = <&timing0>;

            timing0: timing {
                clock-frequency = <30000000>;
                hactive = <800>;
                vactive = <480>;
                hback-porch = <0x5a>;
                hfront-porch = <0x78>;
                vback-porch = <0x1>;
                vfront-porch = <0x1>;
                hsync-len = <0x64>;
                vsync-len = <0x21>;
                //linux,phandle = <0x14>;
                //phandle = &timing0;
                pixelclk-active = "active high";
            };
        };
    };

};

Are you able to tell me more about what that error means? Thanks!