On 2018-04-13 03:15 AM, Michel Dänzer wrote: > On 2018-04-13 04:55 AM, S, Shirish wrote: >> Hi Harry, Alex, >> >> The solution given while reviewing my patch was that "DC should support enabling a CRTC without a framebuffer." > > I think it's weird that an enabled CRTC would end up having a NULL FB > during normal operation in the first place. Any idea how that happens? > I think we've only seen that with ChromeOS and in one case on Ubuntu. If I remember right it was when running GDM with Wayland and logging into X. I imagine our DDX driver never pulls this scenario on us. Technically this would be a reasonable case for our HW, i.e. we could keep pipe running and blanked and just not scan-out a FB. > >> Since the revert is a temporary workaround to address issue at hand and considering the bigger regression it will cause on ChromeOS(explained below), >> I would strongly recommend that the revert should not be mainlined (to linus tree), until a proper fix for both the issues i.e., flickering and BUG hit on atomic is found. >> >> For the sake of everyone's understanding in the list below is a brief background. >> >> Mainline patch from intel folks, "846c7df drm/atomic: Try to preserve the crtc enabled state in drm_atomic_remove_fb, v2." >> introduces a slight behavioral change to rmfb. Instead of disabling a crtc when the primary plane is disabled, it now preserves it. >> >> This change leads to BUG hit while performing atomic commit on amd driver leading to reboot/system instability on ChromeOS which has enabled >> drm atomic way of rendering, I also remember it causing some issue on other OS as well. > > Can you provide more information about "some issue on other OS"? > Otherwise I'm afraid this is a no-brainer, since nobody's using ChromeOS > with an AMD GPU in the wild, whereas many people have been running into > issues with these commits in the wild, especially since they're in the > 4.16 release. > I tend to agree with Michel here, there's been quite the fallout from that patch for people's daily drivers. Harry >
