Re: [PATCH v3 0/2] Initial Exynos850 ACPM support for exynos-acpm

[Krzysztof Kozlowski]

On 29/09/2026 04:11, Alexey Klimov wrote:
> This series adds support for the Exynos850 SoC to the Samsung ACPM
> firmware IPC driver.
> 
> The Exynos850 contains an APM co-processor that handles IPC messages for
> clocks, power, thermal management, and PMIC control. While it uses the
> same underlying ACPM communication machinery and protocols set as
> the GS101, the exact implementation details vary between the two SoCs.
> 
> Those protocol-specific differences will be managed individually by the
> corresponding protocol drivers. However, a dedicated compatible string
> is still required for the core IPC driver because the Exynos850 firmware
> utilizes a different initialisation data base offset.
> 
> First patch updates the google,gs101-acpm-ipc dt-bindings to include the
> new "samsung,exynos850-acpm-ipc" compatible string and adds clocks
> descriptions for clocks that can be controlled via ACPM or by Linux
> directly.
> 
> Second patch adds the corresponding match data, base offset, and clock
> device name to the exynos-acpm firmware driver.
> 
> There is a dependency for driver changes on clk-acpm exynos850 series
> which seems to be accepted/merged:
> https://lore.kernel.org/linux-samsung-soc/20260915-exynos850-acpm-clk-v2-0-ffad4a6b3fb1@xxxxxxxxxx/


I don't understand this dependency. Why does a firmware driver depend on
clock one?

You do understand that mentioning fake dependencies only stalls the
patchset? I would have applied it now, but because of this question, it
will wait and due to conferences might miss this merge window.

Best regards,
Krzysztof