Re: [PATCH v3 0/2] arm64: qcom: sm8550: add DDR, LLCC & L3 CPU bandwidth scaling

[Georgi Djakov]

On 3/10/26 10:31 PM, Aaron Kling wrote:
> On Tue, Mar 10, 2026 at 3:20 PM Krzysztof Kozlowski
> <krzysztof.kozlowski@xxxxxxxxxxxxxxxx> wrote:
> > On 10/03/2026 21:05, Aaron Kling wrote:
> > > > ---
> > > > Aaron Kling (2):
> > > >        dt-bindings: interconnect: OSM L3: Document sm8550 OSM L3 compatible
> > > >        arm64: dts: qcom: sm8550: add cpu OPP table with DDR, LLCC & L3 bandwidths
> > > >
> > > >   .../bindings/interconnect/qcom,osm-l3.yaml         |   1 +
> > > >   arch/arm64/boot/dts/qcom/sm8550.dtsi               | 367 +++++++++++++++++++++
> > > >   2 files changed, 368 insertions(+)
> > > > ---
> > > > base-commit: 9845cf73f7db6094c0d8419d6adb848028f4a921
> > > > change-id: 20260207-sm8550-ddr-bw-scaling-b1524827f207
> > > >
> > > > Best regards,
> > > > --
> > > > Aaron Kling <webgeek1234@xxxxxxxxx>
> > >
> > > What is the normal merge sequence and window for linux-arm-msm? I see
> > > several things that have been picked up for -next recently, but none
> > > of my sm8550 patches that have been reviewed / approved have been
> > > picked up yet.
> >
> >
> > This one is probably waiting on interconnect, no? Not saying that
> > merging here is easy, quite the opposite - it's frustrating, but you can
> > help by responding with actual data, e.g. bindings were merged and DTS
> > can go, instead of just content-less ping.
>
> So patch 1, the bindings, has to go via a different tree; then patch 2
> goes via linux-arm-msm? Or does the first patch need an ack from other
> people? I was assuming both of these could be handled by the
> linux-arm-msm maintainers.
>
> Part of this was a reminder, yes, but the question is still honest. I
> don't know what the expected merge window is here, knowing that is
> good to know if something got lost in the mix. I've got a couple other
> patches as well that are standalone dt changes with no other deps.
> I've had patches to other subsystems that have sat for four or five
> cycles just waiting on the subsystem maintainers.

Hi Aaron,

Last week i picked the 1st patch, so it's in this week's linux-next
releases already. I usually push an immutable branch if there are other
patches that depend on the one i picked, so i did that (icc-sm8550-osm-l3
branch).

Now the Qualcomm maintainers can pick the dts change if they want. My
observations are that the qcom dt tree is closing around -rc5, to give
some time for new patches to get tested before they send a pull request
to the arm-soc maintainers. If some patch is not picked, it's a good
idea to re-base and re-send when the next -rc1 is out.

Thanks,
Georgi