https://bugs.freedesktop.org/show_bug.cgi?id=106597 --- Comment #19 from Lukas Wunner --- Created attachment 139706 --> https://bugs.freedesktop.org/attachment.cgi?id=139706&action=edit drm-next-4.18-wip debug patch 2 Okay at 10.574564, the HDA controller tries to runtime suspend but returns -EBUSY (-16) because the child HDA codec hdaudioC1D0 is still probing. After the child is done with probing at 10.586551, it informs the HDA controller that it may now be idle. However for some reason rpm_idle() doesn't cause the controller to be suspended. Here's another debug patch which also litters the ->runtime_idle() path with debug printks. Please try this and attach the resulting dmesg. > Well, AFAIK the both the HDMI port and the DP on my laptop are directly attached > to the dGPU and not to the iGPU like the internal panel is. Maybe it has > something to do with that? You've got two HDMI codecs in your machine, hdaudioC1D0 on the AMD GPU and hdaudioC0D2 in the Intel PCH. It's unclear to me which one is used for external displays. > Also, as requested: > $ cat /sys/bus/pci/devices/0000:01:00.1/power/runtime_status > active My apologies, I meant runtime_enabled, however it does seem to be enabled, otherwise the idle/suspend machinery wouldn't work. Thanks! -- You are receiving this mail because: You are the assignee for the bug.