Date:   Wed, 10 Apr 2019 12:57:23 +0200
From:   Robert Foss <robert.foss@...labora.com>
To:     Gary Bisson <gary.bisson@...ndarydevices.com>
Cc:     Mark Rutland <mark.rutland@....com>,
        "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
        <devicetree@...r.kernel.org>, Fabio Estevam <festevam@...il.com>,
        Sascha Hauer <s.hauer@...gutronix.de>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        Troy Kisky <troy.kisky@...ndarydevices.com>,
        Rob Herring <robh+dt@...nel.org>,
        NXP Linux Team <linux-imx@....com>,
        Sascha Hauer <kernel@...gutronix.de>,
        Fabio Estevam <fabio.estevam@....com>,
        Shawn Guo <shawnguo@...nel.org>,
        "moderated list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE" 
        <linux-arm-kernel@...ts.infradead.org>
Subject: Re: [PATCH] ARM: dts: imx6qdl-nitrogen6_max: Disable LVDS channels

Hey Gary,

On 4/10/19 9:35 AM, Gary Bisson wrote:
> Hi,
> 
> On Tue, Apr 9, 2019 at 2:07 AM Robert Foss <robert.foss@...labora.com> wrote:
>>
>> Hey Fabio,
>>
>> On 4/8/19 10:37 PM, Fabio Estevam wrote:
>>> Hi Robert,
>>>
>>> [Adding Gary]
> 
> Adding Troy, I'm no longer a full-time employee at Boundary Devices.
> 
>>> On Mon, Apr 8, 2019 at 2:54 PM Robert Foss <robert.foss@...labora.com> wrote:
>>>>
>>>> If a LVDS device is not connected, having the LVDS channels
>>>> enabled will prevent imx-ldb from probing correctly even
>>>> if other CRTCs are connected.
>>>>
>>>> Signed-off-by: Robert Foss <robert.foss@...labora.com>
>>>> ---
>>>>    arch/arm/boot/dts/imx6qdl-nitrogen6_max.dtsi | 4 ++--
>>>>    1 file changed, 2 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/arch/arm/boot/dts/imx6qdl-nitrogen6_max.dtsi b/arch/arm/boot/dts/imx6qdl-nitrogen6_max.dtsi
>>>> index 39200e5dc896..5b413cf4c250 100644
>>>> --- a/arch/arm/boot/dts/imx6qdl-nitrogen6_max.dtsi
>>>> +++ b/arch/arm/boot/dts/imx6qdl-nitrogen6_max.dtsi
>>>> @@ -703,7 +703,7 @@
>>>>           status = "okay";
>>>>
>>>>           lvds-channel@0 {
>>>> -               status = "okay";
>>>> +               status = "disabled";
>>>>
>>>>                   port@4 {
>>>>                           reg = <4>;
>>>> @@ -715,7 +715,7 @@
>>>>           };
>>>>
>>>>           lvds-channel@1 {
>>>> -               status = "okay";
>>>> +               status = "disabled";
>>>
>>> I am not sure I understood what you are trying to fix.
>>
>> If CONFIG_DRM_IMX_LDB is enabled, LVDS DT channels are enabled
>> and no LVDS-panels are connected the imx-ldb driver will
>> fail to bind.
>>
>> This is a problem, since it will prevent other actually connected
>> display output from being used, even if they bind properly.
>>
>>>
>>> Could you please share some logs when imx-ldb fails to probe correctly?
>>
>>
>> [    2.119563] component_bind_all() trying to bind: ldb
>> [    2.124600] imx-drm display-subsystem: binding ldb (ops imx_ldb_ops)
>> [    2.146169] drm_of_find_panel_or_bridge() np->name=lvds-channel  np->type=<NULL>
>> [    2.153709] drm_of_find_panel_or_bridge() no panel found for remote
>> [    2.160081] drm_of_find_panel_or_bridge() bridge needed
>> [    2.162043] drm_of_find_panel_or_bridge() bridge not found
>> [    2.165331] drm_of_find_panel_or_bridge() failed
>> [    2.170023] imx-drm display-subsystem: failed to bind ldb (ops imx_ldb_ops): -517
>>
>> This at the same time as HDMI binds properly:
>>
>> [    4.458954] dwhdmi-imx 120000.hdmi: Detected HDMI TX controller v1.30a with
>> HDCP (DWC HDMI 3D TX PHY)
>> [    4.469633] imx-drm display-subsystem: bound 120000.hdmi (ops dw_hdmi_imx_ops)
>>
>> Which in the end causes the IMX driver to not initialize properly
>> and ignore the HDMI connection that bound properly.
>> This in turn prevents us from having any graphical output while there
>> is no LVDS panel connected.
> 
> Does this happen if the LVDS is disabled in cmdline? (adding "video=LVDS-1:d")

I hadn't tried this before, but is seems yo yield the same results

# cat /proc/cmdline
video=LVDS-1:d enforcing=0 ip=dhcp rw rootwait root=/dev/mmcblk1p1 
log_buf_len=16M cma=512M vmalloc=512M

# dmesg
imx-drm display-subsystem: failed to bind disp0 (ops imx_pd_ops): -517
imx-drm display-subsystem: master bind failed: -517

This dmesg log is failing to bind disp0 at another point than what I
was seeing before, yet the result is the same. The connected HDMI
output is not brought online correctly.

> 
> Note that a bootscript was written for mainline kernel that takes care
> of display configuration [1].
> 
> Currently this bootscript disables all the displays that aren't used.
> It was tested with HDMI, where LVDS and LCD display are disabled, and
> it *used to* work.
> 
> If this doesn't work then there's most likely a regression somewhere.
> 
> Regards,
> Gary
> 
> [1] https://github.com/boundarydevices/u-boot-imx6/blob/boundary-v2018.07/board/boundary/bootscripts/bootscript-mainline.txt#L90
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@...ts.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> 

Powered by blists - more mailing lists