Hi Sylwester,
On 13.11.2020 11:32, Sylwester Nawrocki wrote:
> On 13.11.2020 10:07, Chanwoo Choi wrote:
>> On 11/13/20 5:48 PM, Georgi Djakov wrote:
>>> On 11/12/20 16:09, Sylwester Nawrocki wrote:
> [...]
>>> Good work Sylwester! Thank you and all the reviewers! What would be the merge
>>> path for this patchset? Looks like there is no build dependency between patches.
>>> Should i take just patches 2,3 or also patch 1? Chanwoo?
>> Hi Georgi,
>>
>> If you take the patch 2,3, I'll apply patch 1,4 to devfreq.git.
>> Hi Sylwester,
>> First of all, thanks for your work to finish it for a long time.
>> I'm very happy about finishing this work. It is very necessary feature
>> for the QoS. Once again, thank for your work.
> I would also like to thank everyone for provided feedback!
>
> As far as building is concerned the patches could be applied in any
> order. I think we could also apply the drm/exynos patch in same
> merge window. There could be runtime (or git bisect) regression
> only in case when INTERCONNECT is enabled and only (or as first)
> the dts and drm/exynos patches are applied.
>
> Hmm, maybe it's better to hold on with the drm patch, INTERCONNECT
> is disabled in arch/arm/configs/{multi_v7_defconfig, exynos_defconfig}
> but it is enabled in arch/arm64/configs/defconfig.
I don't think we need to delay DRM patch. Exynos DRM mixer is not
available on ARM64 SoCs, so this won't be an issue.
Best regards
--
Marek Szyprowski, PhD
Samsung R&D Institute Poland
