15:30 <tolszak> narmstrong: I'm sorry but seems like the bug I found may be in meson drm and looking in kernel MAINTAINERS list you are the meson drm maintainer. What data would be the best to narrow down the issue? I have apitrace from egl but seems like it will not be helpful.

15:31 <tolszak> I will check if issue exists in plain almogic image sato. It happens on my custom distro, with changes in kernel config (adding a few more usb devices to handle)

15:33 <narmstrong> tolszak: please be more explicit, the meson DRM driver is only here to display the buffers generated by the GPU

15:34 <tolszak> narmstrong: Im sorry for not being precise, I've tested example app and with LIBGL_ALWAYS_SOFTWARE=1

15:34 <tolszak> Hence on panfrost channel I was told it must be meson drm dirver issue

15:35 <narmstrong> indeed if LIBGL_ALWAYS_SOFTWARE=1 panfrost is no more involved

15:36 <narmstrong> but I don't see how a GL shader parameter can break the display output, or I haven't understood the issue correctly

15:36 <tolszak> narmstrong: I already asked, but it is enough to just run the opengl application even if GLES library is loaded implicitly right?

15:36 <tolszak> narmstrong: I mean the app doesn't need to link it

15:37 <tolszak> macc24: "it's then *checks backlog* meson_drm bug"

15:37 <tolszak> that's the output of the discussion

15:39 <narmstrong> I still don't understand the issue...

15:41 <tolszak> I will grab a few pictures and send you over email, or a video

15:52 <tolszak> narmstrong: Sent together with follow up email which explains attached file names.