Re: [PATCH v5 5/8] clk: qcom: gcc-msm8939: mark Venus core GDSCs as hardware controlled

On 09/05/2026 15:43, Erikas Bitovtas wrote:
>> OTOH if Core1 can't be coaxed into working in either mode, we are better
>> off either holding off on this series until the breakages is root caused
>> or leaving out Core1 entirely for now.
>>
>> i.e. there is no utility in declaring an encoder or decoder to userspace
>> that is known to be broken.
>> Last time I tried to enable only 1 of the cores, the following would happen:
> - If core0 is enabled, HEVC decoding would work, but not the other codecs.
> - If core1 is enabled, the rest of codecs would work, but not HEVC.

Can you be a little clearer here do you mean - doesn't work on that core 
or doesn't work on all cores ?

> Given we don't really have a way to power up a core based on which codec
> we are dealing with, I suggest we leave both cores in the binding and
> enable them indiscriminately.

OK the 8916 to 8939 migration guide says HEVC decoder - encoder is not 
mentioned.

You can have a bitmask flag of whitelist or blacklist per core then - 
select which codecs are supported on which cores and then declare those 
to user-space.

---
bod