| Summary: | amd-staging-drm-next, R4 Mullins hdmi monitor wont wake up from S3 - *ERROR* REG_WAIT timeout 1us * 80000 tries - dce_abm_immediate_disable | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | DRI | Reporter: | Przemek <soprwa> | ||||||
| Component: | DRM/AMDgpu | Assignee: | Default DRI bug account <dri-devel> | ||||||
| Status: | RESOLVED FIXED | QA Contact: | |||||||
| Severity: | normal | ||||||||
| Priority: | medium | CC: | FD, harry.wentland, me, nicholas.kazlauskas, sunpeng.li | ||||||
| Version: | DRI git | ||||||||
| Hardware: | x86-64 (AMD64) | ||||||||
| OS: | Linux (All) | ||||||||
| Whiteboard: | |||||||||
| i915 platform: | i915 features: | ||||||||
| Attachments: |
|
||||||||
|
Description
Przemek
2018-10-27 10:29:13 UTC
Created attachment 142606 [details]
picture of the screen after resume
With latest (yesterday's) amd-staging-drm-next repo pull situation gets worse.
After fresh boot and when drm amdgpu fb gets initialized laptop internal screen gets off and picture appears on external monitor connected to hdmi output.
After sddm login, when plasma desktop fires up picture appears on eDP screen and both (hdmi and internal) are working properly until suspending (s3).
After resume from suspend there is no picture on hdmi monitor, brightess level is reset to 0, on eDP picture is shuttered and full of glitches as you can see on attached photo (it does no matter if it is xorg or tty, both have those glitches).
Thanks,
Przemek.
Created attachment 142828 [details]
log with drm.debug=0x4 after wake up from suspend
Glitched screen issue has been resolved (I think - Thank you very much), but the issue with black screens still persist in amd-staging-drm-next repo synced on 15th of December.
There is a strange output in attached syslog, amdgpu is saying that hdmi monitor is disconnected while it's not true. It is connected and powered on, but its look like connector doesn't see it.
Thanks,
Przemek.
It seems also, that WARNING from the first comment is not related to this situation, it occurs, but less often tough, so in this case should we change the title of this bug report or leave it as is? Thanks, Przemek. Yesterday I have observed a strange behavior. After suspend/resume cycle HDMI screen is turned off, but when you trigger hibernate/resume after suspend/resume HDMI output is turned back on and is working properly. (Yes, hibernation works on CIK and amdgpu now!! Thank you so much!!) As given above a question arise, are there other-different triggers/resets in suspend/resume and hibernate/resume cycles for those asics? Thanks, Przemek. As for today's git pull everything (suspend/hibernate/resume) seems to work just fine. I have connected SAMSUNG TV to HDMI output and it wakes up properly. The problem occurs only on hdmi -> vga display adapter which I'm using to connect my old PHILIPS 170C lcd monitor. (VGA connector is not implemented on amdgpu). I'm closing this bug report as FIXED, but still wondering why adapter has to be replugged to make monitor wake up/being discovered by amdgpu driver, tough. Thank you very much. Przemek |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.