Re: [PATCH RFC/WIP 0/4] arm64: dts: qcom: sm8750: Enable display

[Jessica Zhang]

On 10/28/2025 11:27 PM, Krzysztof Kozlowski wrote:
> On 29/10/2025 07:20, Mahadevan P wrote:
> > Hi Krzysztof,
> >
> > On 4/26/2025 1:24 AM, Krzysztof Kozlowski wrote:
> > > On 25/04/2025 21:34, Dmitry Baryshkov wrote:
> > > > On Thu, Apr 24, 2025 at 03:04:24PM +0200, Krzysztof Kozlowski wrote:
> > > > > DTS is ready and I consider it ready for review, but still RFC because:
> > > > > 1. Display has unresolved issues which might result in change in
> > > > >      bindings (clock parents),
> > > > > 2. I did not test it since some time on my board...
> > > > > 3. Just want to share it fast to unblock any dependent work.
> > > > >
> > > > > DTS build dependencies - as in b4 deps, so:
> > > > > https://lore.kernel.org/r/20250421-sm8750_usb_master-v5-0-25c79ed01d02@xxxxxxxxxxxxxxxx/
> > > > > https://lore.kernel.org/r/20250424-sm8750-audio-part-2-v1-0-50133a0ec35f@xxxxxxxxxx/
> > > > > https://lore.kernel.org/r/20250113-sm8750_gpmic_master-v1-2-ef45cf206979@xxxxxxxxxxx/
> > > > >
> > > > > Bindings:
> > > > > 1. Panel:https://github.com/krzk/linux/tree/b4/sm8750-display-panel
> > > > > 2. MDSS:https://lore.kernel.org/r/20250311-b4-sm8750-display-v4-0-da6b3e959c76@xxxxxxxxxx/
> > > > >
> > > > > Patchset based on next-20250424.
> > > > If the DisplayPort works on this platform, I'd kindly ask to send
> > > > separate DP+DPU only series (both driver and arm64/dts). It would make
> > > > it much easier (at least for me) to land the series, while you and
> > > > Qualcomm engineers are working on the DSI issues.
> > > DP has also issues - link training failures, although it feels as
> > > different one, because DSI issue Jessica narrowed to DSI PHY PLL VCO
> > > rate and I have a working display (just don't know how to actually solve
> > > this).
> >
> > We at Qualcomm are currently working on bringing up the DSI display on
> > MTP. For this, I’ve picked the following patches on top of |v6.18-rc2|:
>
> Display on MTP8750 was fully brought already. I don't understand why you
> are doing the same.
>
> >   1. All the DT changes mentioned in this series
> >   2. [PATCH v2] drm/msm/dpu: Fix adjusted mode clock check for 3d merge
> >      https://lore.kernel.org/all/1154f275-f934-46ae-950a-209d31463525@xxxxxxxxxx/
> >   3. [PATCH v2 0/2] drm/panel: Add Novatek NT37801 panel driver
> >      https://lore.kernel.org/all/20250508-sm8750-display-panel-v2-0-3ca072e3d1fa@xxxxxxxxxx/
> >
> > However, when testing with |modetest|, the panel appears blank. I wanted
> > to check if there are any additional patches already posted that I might
> > have missed and should be included.
>
> Many patches are missing because Qualcomm did not send them for months.
> I was pinging multiple times and I gave up - my job is not ping Qualcomm
> to send their patches.
>
> I have no clue what you have already, what is your base, what errors you
> have and I will not be guessing this. For convenience, you can use my
> integration WIP branch from my github. I already shared that tree with
> Qualcomm more than once. Please reach internally first, instead of
> poking community.
>
>
> > Also, I’m curious to understand more about the DSI PHY PLL VCO rate
> > issue that Jessica had narrowed down—could you please share some details?
>
> Sorry, I am not going to repeat stuff done year ago. Please reach to the
> archives.

I don't have my notes on me, but IIRC this was the byte/pixel clk RCG 
failed to update issue that was causing the clocks to default to a bad 
rate (despite driver requesting the correct rate). It was fixed by this 
patch [1].

Thanks,

Jessica Zhang

[1] 
https://patchwork.kernel.org/project/linux-arm-msm/patch/20250520090741.45820-2-krzysztof.kozlowski@xxxxxxxxxx/

> > Lastly, I’d appreciate it if you could share the plan for merging these
> > changes upstream.
> I don't understand what you ask me. Process of contributing to Linux
> kernel is well documented.
>
> Best regards,
> Krzysztof