On 1/8/2025 12:41 PM, Bjorn Andersson wrote:
On Thu, Jan 02, 2025 at 12:06:14PM +0800, Renjiang Han wrote:
On 12/26/2024 11:54 AM, Bjorn Andersson wrote:
The video driver will be using the newly introduced
dev_pm_genpd_set_hwmode() API to switch the video GDSC to HW and SW
control modes at runtime.
"Will be using", does that imply then that if this patch is merged
before (or without the venus patch) something unexpected will happen?
Please clarify how you expect this to be merged, or clarify in the
commit message that ordering is not of any concern.
Regards,
Bjorn
Thanks for your comment. This patch series is to make the video driver
to use dev_pm_genpd_set_hwmode() to switch GDSC mode. This patch and
the venus driver patch need to be merged at the same time. Otherwise,
the video will not work properly on these platforms.
The two patches are handled by different maintainers, of different
subsystems and as such would not be expected to be merged together ever.
If you have such requirements, it need to be made very clear to the
maintainers that they will have to synchronize the effort.
You're expected to always keep the tree "bisectable", i.e. the tree
should function after each commit in the git history. Please clarify
the best possible order here, and if the changes truly need to be merged
in some specific order let's see if we can get Maruo's Acked-by and I
could merge the pair through the clock tree.
Regards,
Bjorn
Thanks for your explanation. The use of dev_pm_genpd_set_hwmode() depends
on the HW_CTRL_TRIGGER flag, and the reading and writing of the
WRAPPER_VCODEC0_MMCC_POWER_CONTROL and WRAPPER_VCODEC0_MMCC_POWER_STATUS
registers depends on the HW_CTRL flag. Therefore, the clock patch and
the venus driver patch need to be merged at the same time. Otherwise,
the venus driver cannot work properly.
--
Best Regards,
Renjiang
