> I can drop the patch if you feel strongly, although it is pretty > similar your original patch, it just keeps all of the logic within > navi10_ppt.c which is cleaner in my opinion rather than pulling navi1x > specific logic into the core pm code. > > Alex
I dont have problem with the code, the problem is that, this is not the solution to this problem, as i said in the previous, the solution lies in the DC not the SMU (i tested by removing dce_v10_0_early_init and adding that function to dce_v10_sw_init), it is also confirmed via the testing using Opencore Lecacy patcher, which makes the display 5K, and make the display not use compatibility mode (4K), and it also fixes the SMU problem. (https://gitlab.freedesktop.org/drm/amd/-/work_items/4455) You can choose to upstream this, i dont necessarily have a problem with this patch as long as i am the patch author (eg using From: ) Thanks
