00:15 <lumag> ektor5: what kind of artifacts? Could you possibly capture a video?

00:15 <lumag> abhinav__: ^^

00:17 <ektor5> lumag: ofc, would it be better to send it here or in msm gitlab issue tracker?

00:24 <ektor5[m]> Here a screenshot from Superhot MCR, seems that the right half is not understanding some "overlay" correctly, most of the gameplay is fine, but I see problems with things like shadows and menus

00:28 <ektor5[m]> Running on jhovold wip/x1e80100-6.15-rc6

00:32 <ektor5[m]> Civilization IV, benchmark mode

00:36 <ektor5> s/IV/VI/

11:02 <vknecht[m]> <bryanodonoghue> "vknecht so that register sets..." <- Hi ! do the last two commits here look okay to you ? https://github.com/vknecht/msm8916-mainline-linux/commits/msm8939-camss-upd

16:36 <narmstrong> lumag: looks like the alpha bits gets interpreted on the right side while not on the left side even if the format is xr24, like if the format config of the right side of the parralel multirect is incorrect

16:37 <lumag> narmstrong: thanks, I will take a look

16:38 <lumag> ektor5: could you possibly dump /sys/kernel/debug/dri/0/state at this point?

16:39 <ektor5> lumag: yes, I'll send you good and bad state

16:39 <lumag> thanks!

17:01 <abhinav__> narmstrong lumag that might be possible, but usually the blend issue should affect the whole layer not selectively blend incorrectly. I agree, we should get a state dump in this state and take a look.

17:08 <lumag> ektor5: could you please start an issue on gitlab.fdo/drm/msm and put the screenshot, video and state dump there?

17:09 <ektor5> lumag: yes absolutely