arm64: dts: qcom: common dtsi for Xiaomi SDM660-family (upstream prep)

[M0Rf30]

Per discussion in #161 and #165, this prepares the common dtsi for upstream submission.

Based on torvalds/master (mainline).

Patchset:

1. Create sdm660-xiaomi-common.dtsi, refactor lavender to use it
2. dt-bindings for tulip, whyred, jasmine, platina
3. Add tulip DTS (Redmi Note 6 Pro)
4. Add whyred DTS (Redmi Note 5 Pro)
5. Add jasmine DTS (Mi A2)
6. Add platina DTS (Mi 8 Lite)

Validation:

• checkpatch.pl --strict: 0 errors
• make CHECK_DTBS=y: passed
• All DTBs build

Please review before I send to linux-arm-msm@vger.kernel.org.

[setotau]

there are over 10000 commits, how can we review these changes?

[setotau]

there are over 10000 commits, how can we review these changes?

nvm, there are last commits I see

[okias]

I'll drop some nitpicks:

chosen > framebuffer - use

memory-region = <&framebufer_mem>;

instead of reg =

Except this small improvements, looks pretty good.

[setotau]

lgtm

[okias]

LGTM, last nitpick:
arm64: dts: qcom: add sdm660-xiaomi-common.dtsi for Xiaomi SDM660-family devices

should more like rather than "add" say something like "extract common parts into xiaomi-common". you can omit .dtsi in the commit msg

[minlexx]

And I think SoB chain is wrong in most commit messages where several authors are involved (missing co-developed-by?)

At least commit messages for whyred and platina

[minlexx]

And I think SoB chain is wrong in most commit messages where several authors are involved (missing co-developed-by?)

At least commit messages for whyred and platina

I think in commit messages SoB chain should go top-to-bottom in chronological order: https://docs.kernel.org/process/submitting-patches.html#when-to-use-acked-by-cc-and-co-developed-by

So, for whyred:

Co-developed-by: Mathesh Velayudan <123v.mathesh@gmail.com>
Signed-off-by: Mathesh Velayudan <123v.mathesh@gmail.com>
Co-developed-by: Alexey Minnekhanov <alexeymin@postmarketos.org>
Signed-off-by: Alexey Minnekhanov <alexeymin@postmarketos.org>
Signed-off-by: Gianluca Boiano <morf3089@gmail.com>

for platina also Angelo was the original author, so:

Co-developed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
Co-developed-by: Molly Sophia <mollysophia379@gmail.com>
Signed-off-by: Molly Sophia <mollysophia379@gmail.com>
Signed-off-by: Gianluca Boiano <morf3089@gmail.com>

[minlexx]

Also your every device's commit message says "..USB support for debugging." but it's a normal USB, not just for debugging? 😺

[minlexx]

I am thinking about maybe another base could be made sdm660-pm660.dtsi for all devices what we have. This approach what msm8916-mainline developers have done.

https://github.com/msm8916-mainline/linux/blob/msm8916/6.17-rc6/arch/arm64/boot/dts/qcom/msm8916-pm8916.dtsi

And then what to put in it? It will be question every time. I think it's very hard to justify its existence for 3 nodes - it doesn't do much. I think it was much better in the first revision where common regulators were in sdm660-xiaomi-common (sorry!). Also I think it's better to keep this patchset limited to only xiaomi devices and not touch any other Inforce or Sony boards. Thing is, regulators setup is not final yet and will require more patches (like xiaomi-specific setting l3b to always-on) and maybe some more (raising voltage for SD stability one). Maybe such optimisation can be postponed for later, and for now better to keep this patchset as simple as possible?

IMO previous iteration was fine except missing pinctrl for gpio keys and second major issue is missing touchscreen support for boards, but touch can be added later I think and is not related to having common dtsi 🤔 .

Let's decide where to put xiaomi regulators (@M0Rf30 if you think it's possible to have them all in common then OK).

[minlexx]

Should be fine for first revision. I doubt we can find more issues here (but upstream probably can). Instead of arguing here for more days maybe it's time to move the rest questionable subjects to upstream discussion and send it. Of course, if patcheset compiles fine when applied on top of current torvalds/master, checkpatch reports no warnings and dtbs-check also :)

P.S> Sadly our CI doesn't run on master branch, I was thinking of adding checkpatch step to it also.

[okias]

Pretty good, last request: sort the nodes. After this I'm happy to put R-b on the series (it's better to do after it's send, just add me into

Cc: David Heidelberg <david@ixit.cz>

[M0Rf30]

I am thinking about maybe another base could be made sdm660-pm660.dtsi for all devices what we have. This approach what msm8916-mainline developers have done.
https://github.com/msm8916-mainline/linux/blob/msm8916/6.17-rc6/arch/arm64/boot/dts/qcom/msm8916-pm8916.dtsi

And then what to put in it? It will be question every time. I think it's very hard to justify its existence for 3 nodes - it doesn't do much. I think it was much better in the first revision where common regulators were in sdm660-xiaomi-common (sorry!). Also I think it's better to keep this patchset limited to only xiaomi devices and not touch any other Inforce or Sony boards. Thing is, regulators setup is not final yet and will require more patches (like xiaomi-specific setting l3b to always-on) and maybe some more (raising voltage for SD stability one). Maybe such optimisation can be postponed for later, and for now better to keep this patchset as simple as possible?

IMO previous iteration was fine except missing pinctrl for gpio keys and second major issue is missing touchscreen support for boards, but touch can be added later I think and is not related to having common dtsi 🤔 .

Let's decide where to put xiaomi regulators (@M0Rf30 if you think it's possible to have them all in common then OK).

I leave regulators there (only 3 of them are common).
I added touchscreen where possible:

[barni2000]

Send it! =)

[M0Rf30]

sent
https://lore.kernel.org/linux-arm-msm/20260120180052.1031231-1-morf3089@gmail.com/

🤞

[okias]

Nice! If there will be next revision, you can throw in

Cc: phone-devel@vger.kernel.org

into cover letter.

[minlexx]

This can be closed now, because local review is finished and series sent?