Re: [PATCH v2 2/5] arm64: dts: imx8qxp: Add display controller subsystem

From: Alexander Stein

Date: Tue Oct 20 2026 - 03:16:39 EST

Hi,

Am Dienstag, 20. Oktober 2026, 03:47:21 Mitteleuropäische Sommerzeit schrieb Liu Ying:
> On Fri, Oct 16, 2026 at 02:46:47PM +0200, Alexander Stein wrote:
> > Hi,
>
> Hi Alexander,
>
> >
> > Am Donnerstag, 15. Oktober 2026, 05:41:54 Mitteleuropäische Sommerzeit schrieb Liu Ying:
> >> Hi Alexander,
> >>
> >> On Wed, Oct 14, 2026 at 09:16:46AM +0200, Alexander Stein wrote:
> >>
> >> [...]
> >>
> >>> + dc0: display-controller@56180000 {
> >>> + reg = <0x56180000 0x40000>;
> >>> + clocks = <&dc0_lpcg IMX_LPCG_CLK_4>;
> >>> + power-domains = <&pd IMX_SC_R_DC_0>;
> >>> + #address-cells = <1>;
> >>> + #size-cells = <1>;
> >>> + ranges;
> >>> + status = "disabled";
> >>> +
> >>> + dc0_intc: interrupt-controller@56180040 {
> >>> + reg = <0x56180040 0x60>;
> >>> + clocks = <&dc0_lpcg IMX_LPCG_CLK_5>;
> >>> + interrupt-controller;
> >>> + interrupt-parent = <&dc0_irqsteer>;
> >>> + #interrupt-cells = <1>;
> >>> + interrupts = <448>, <449>, <450>, <64>,
> >>> + <65>, <66>, <67>, <68>,
> >>> + <69>, <70>, <193>, <194>,
> >>> + <195>, <196>, <197>, <72>,
> >>> + <73>, <74>, <75>, <76>,
> >>> + <77>, <78>, <79>, <80>,
> >>> + <81>, <199>, <200>, <201>,
> >>> + <202>, <203>, <204>, <205>,
> >>> + <206>, <207>, <208>, <5>,
> >>> + <0>, <1>, <2>, <3>,
> >>> + <4>, <82>, <83>, <84>,
> >>> + <85>, <209>, <210>, <211>,
> >>> + <212>;
> >>> + interrupt-names = "store9_shdload",
> >>> + "store9_framecomplete",
> >>> + "store9_seqcomplete",
> >>> + "extdst0_shdload",
> >>> + "extdst0_framecomplete",
> >>> + "extdst0_seqcomplete",
> >>> + "extdst4_shdload",
> >>> + "extdst4_framecomplete",
> >>> + "extdst4_seqcomplete",
> >>> + "extdst1_shdload",
> >>> + "extdst1_framecomplete",
> >>> + "extdst1_seqcomplete",
> >>> + "extdst5_shdload",
> >>> + "extdst5_framecomplete",
> >>> + "extdst5_seqcomplete",
> >>> + "disengcfg_shdload0",
> >>> + "disengcfg_framecomplete0",
> >>> + "disengcfg_seqcomplete0",
> >>> + "framegen0_int0",
> >>> + "framegen0_int1",
> >>> + "framegen0_int2",
> >>> + "framegen0_int3",
> >>> + "sig0_shdload",
> >>> + "sig0_valid",
> >>> + "sig0_error",
> >>> + "disengcfg_shdload1",
> >>> + "disengcfg_framecomplete1",
> >>> + "disengcfg_seqcomplete1",
> >>> + "framegen1_int0",
> >>> + "framegen1_int1",
> >>> + "framegen1_int2",
> >>> + "framegen1_int3",
> >>> + "sig1_shdload",
> >>> + "sig1_valid",
> >>> + "sig1_error",
> >>> + "reserved",
> >>> + "cmdseq_error",
> >>> + "comctrl_sw0",
> >>> + "comctrl_sw1",
> >>> + "comctrl_sw2",
> >>> + "comctrl_sw3",
> >>> + "framegen0_primsync_on",
> >>> + "framegen0_primsync_off",
> >>> + "framegen0_secsync_on",
> >>> + "framegen0_secsync_off",
> >>> + "framegen1_primsync_on",
> >>> + "framegen1_primsync_off",
> >>> + "framegen1_secsync_on",
> >>> + "framegen1_secsync_off";
> >>> + };
> >>> +
> >>> + dc0_pixel_engine: pixel-engine@56180800 {
> >>> + reg = <0x56180800 0xac00>;
> >>> + clocks = <&dc0_lpcg IMX_LPCG_CLK_5>;
> >>> + #address-cells = <1>;
> >>> + #size-cells = <1>;
> >>> + ranges;
> >>> +
> >>> + dc0_constframe0: constframe@56180960 {
> >>> + reg = <0x56180960 0xc>, <0x56184400 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_extdst0: extdst@56180980 {
> >>> + reg = <0x56180980 0x1c>, <0x56184800 0x28>;
> >>> + reg-names = "pec", "cfg";
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <3>, <4>, <5>;
> >>> + interrupt-names = "shdload", "framecomplete", "seqcomplete";
> >>> + };
> >>> +
> >>> + dc0_constframe4: constframe@561809a0 {
> >>> + reg = <0x561809a0 0xc>, <0x56184c00 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_extdst4: extdst@561809c0 {
> >>> + reg = <0x561809c0 0x1c>, <0x56185000 0x28>;
> >>> + reg-names = "pec", "cfg";
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <6>, <7>, <8>;
> >>> + interrupt-names = "shdload", "framecomplete", "seqcomplete";
> >>> + };
> >>> +
> >>> + dc0_constframe1: constframe@561809e0 {
> >>> + reg = <0x561809e0 0xc>, <0x56185400 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_extdst1: extdst@56180a00 {
> >>> + reg = <0x56180a00 0x1c>, <0x56185800 0x28>;
> >>> + reg-names = "pec", "cfg";
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <9>, <10>, <11>;
> >>> + interrupt-names = "shdload", "framecomplete", "seqcomplete";
> >>> + };
> >>> +
> >>> + dc0_constframe5: constframe@56180a20 {
> >>> + reg = <0x56180a20 0xc>, <0x56185c00 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_extdst5: extdst@56180a40 {
> >>> + reg = <0x56180a40 0x1c>, <0x56186000 0x28>;
> >>> + reg-names = "pec", "cfg";
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <12>, <13>, <14>;
> >>> + interrupt-names = "shdload", "framecomplete", "seqcomplete";
> >>> + };
> >>> +
> >>> + dc0_fetchwarp2: fetchwarp@56180a60 {
> >>> + reg = <0x56180a60 0x10>, <0x56186400 0x190>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_fetchlayer0: fetchlayer@56180ac0 {
> >>> + reg = <0x56180ac0 0xc>, <0x56188400 0x404>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_layerblend0: layerblend@56180ba0 {
> >>> + reg = <0x56180ba0 0x10>, <0x5618a400 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_layerblend1: layerblend@56180bc0 {
> >>> + reg = <0x56180bc0 0x10>, <0x5618a800 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_layerblend2: layerblend@56180be0 {
> >>> + reg = <0x56180be0 0x10>, <0x5618ac00 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> +
> >>> + dc0_layerblend3: layerblend@56180c00 {
> >>> + reg = <0x56180c00 0x10>, <0x5618b000 0x20>;
> >>> + reg-names = "pec", "cfg";
> >>> + };
> >>> + };
> >>> +
> >>> + dc0_display_engine0: display-engine@5618b400 {
> >>> + reg = <0x5618b400 0x14>, <0x5618b800 0x1c00>;
> >>> + reg-names = "top", "cfg";
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <15>, <16>, <17>;
> >>> + interrupt-names = "shdload", "framecomplete", "seqcomplete";
> >>> + power-domains = <&pd IMX_SC_R_DC_0_PLL_0>;
> >>> + #address-cells = <1>;
> >>> + #size-cells = <1>;
> >>> + ranges;
> >>> +
> >>> + dc0_framegen0: framegen@5618b800 {
> >>> + reg = <0x5618b800 0x98>;
> >>> + clocks = <&dc0_disp_lpcg IMX_LPCG_CLK_0>;
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <18>, <19>, <20>, <21>,
> >>> + <41>, <42>, <43>, <44>;
> >>> + interrupt-names = "int0", "int1", "int2", "int3",
> >>> + "primsync_on", "primsync_off",
> >>> + "secsync_on", "secsync_off";
> >>> + };
> >>> +
> >>> + dc0_tcon0: tcon@5618c800 {
> >>> + reg = <0x5618c800 0x588>;
> >>> +
> >>> + port {
> >>> + dc0_disp0_dc0_pixel_combiner_ch0: endpoint {
> >>> + remote-endpoint = <&dc0_pixel_combiner_ch0_dc0_disp0>;
> >>> + };
> >>> + };
> >>> + };
> >>> + };
> >>> +
> >>> + dc0_display_engine1: display-engine@5618b420 {
> >>> + reg = <0x5618b420 0x14>, <0x5618d400 0x1c00>;
> >>> + reg-names = "top", "cfg";
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <25>, <26>, <27>;
> >>> + interrupt-names = "shdload", "framecomplete", "seqcomplete";
> >>> + power-domains = <&pd IMX_SC_R_DC_0_PLL_1>;
> >>> + #address-cells = <1>;
> >>> + #size-cells = <1>;
> >>> + ranges;
> >>> +
> >>> + dc0_framegen1: framegen@5618d400 {
> >>> + reg = <0x5618d400 0x98>;
> >>> + clocks = <&dc0_disp_lpcg IMX_LPCG_CLK_1>;
> >>> + interrupt-parent = <&dc0_intc>;
> >>> + interrupts = <28>, <29>, <30>, <31>,
> >>> + <45>, <46>, <47>, <48>;
> >>> + interrupt-names = "int0", "int1", "int2", "int3",
> >>> + "primsync_on", "primsync_off",
> >>> + "secsync_on", "secsync_off";
> >>> + };
> >>> +
> >>> + dc0_tcon1: tcon@5618e400 {
> >>> + reg = <0x5618e400 0x588>;
> >>> +
> >>> + port {
> >>> + dc0_disp1_dc0_pixel_combiner_ch1: endpoint {
> >>> + remote-endpoint = <&dc0_pixel_combiner_ch1_dc0_disp1>;
> >>> + };
> >>> + };
> >>> + };
> >>> + };
> >>> + };
> >>
> >> I'm thinking that it would be good to have a complete display controller
> >> DT node in the first place, including those prefetch engines attached to
> >> fetchunits through the fsl,prg property. In this way, display controller
> >> driver's life would be easier - all child/grandchild devices are always
> >> available. I'm considering to adding a check to the Linux display controller
> >> driver to ensure the availability(at least for critical child/grandchild
> >> devices).
> >
> > Well, this sounds reasonable, but this is beyond my knowledge about the
> > internals of the display controller.
> > FWIW it works (at least in my case), any additional DT node could be added
> > later on.
>
> If someone says that new kernel must support old device tree, then the display
> driver's life would be not be easy.

Are the upcoming DT changes incompatible? Well then let's get them merged
first.

> > If there is something to test/review feel free to send it onto the
> > list. Please keep me on CC.
>
> I can try to CC you, but in case I forget to do so, the proper way for you
> is to get any patch in question by yourself - lei + mutt is a typical solution.

Oh nice. I wasn't aware of lei. Thanks for the hint.

Best regards
Alexander
--
TQ-Systems GmbH | Mühlstraße 2, Gut Delling | 82229 Seefeld, Germany
Amtsgericht München, HRB 105018
Geschäftsführer: Detlef Schneider, Rüdiger Stahl, Stefan Schneider
http://www.tq-group.com/

Attachment: signature.asc
Description: This is a digitally signed message part.