Summary: I'm seeing a very brief screen flicker, uncertain of the interval (perhaps once per half hour), whether using built-in only display or external (appears on both displays). This is a regression, it doesn't happen with 4.18.x. model name : Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz 00:02.0 VGA compatible controller: Intel Corporation Skylake GT2 [HD Graphics 520] (rev 07) I've enabled drm.debug=0x1e but can only isolate to about 1 minute (I see the flicker, look at the time, and then use journalctl --since= to capture from the start of that minute.
Created attachment 142279 [details] journal snippet A less than 1 minute snippet from kernel messages with drm.debug=0x1e enabled, at the time the flicker happens.
Created attachment 142280 [details] journal full, kmesg only
Please note the 2nd attachment, full journal, is for a subsequent boot from the journal snippet attachment. The /run/ buffer ran out of space (days of uptime) and had started deleting journal entries, so I did a clean boot to get a dmesg that includes all the early boot stuff.
Chris, Can you say what is the usecase here? Any particular pattern that causes this issue?
Fedora 29 Workstation, laptop. Sometimes it happens in web browser (Firefox) sometimes in Terminal. I haven't found a pattern.
(In reply to Chris Murphy from comment #5) > Fedora 29 Workstation, laptop. Sometimes it happens in web browser (Firefox) > sometimes in Terminal. I haven't found a pattern. Can you attach cat /sys/kernel/debug/dri/0/i915_edp_psr.
There are only two edp_psr items: [root@flap ~]# cat /sys/kernel/debug/dri/0/i915_edp_psr_debug cat: /sys/kernel/debug/dri/0/i915_edp_psr_debug: Operation not permitted [root@flap ~]# cat /sys/kernel/debug/dri/0/i915_edp_psr_status Sink_Support: yes Enabled: no Busy frontbuffer bits: 0x000 Main link in standby mode: no HW Enabled & Active bit: no Source PSR status: 0x4000000 [IDLE]
Can you attach dmesg from boot with kernel parameters drm.debug=0x1e log_buf_len=4M. Have you verified this issue with latest drm-tip?(https://cgit.freedesktop.org/drm-tip)
>Can you attach dmesg from boot with kernel parameters drm.debug=0x1e Both attachments have drm.debug=0x1e enabled. >Have you verified this issue with latest drm-tip? No, I can't build kernels while I'm mobile, but I can install a newer Fedora Rawhide kernel if that's useful.
> No, I can't build kernels while I'm mobile, but I can install a newer Fedora > Rawhide kernel if that's useful. Yes, dmesg from boot with latest stable kernel will help in this case. Ensure, kernel parameters are set (drm.debug=0x1e log_buf_len=4M)
The two attachments are based on 4.19.0; Fedora is in the middle of rebasing to 4.19 so for whatever reason there won't be any 4.19 builds until that happens. But they have 4.20rc1 built which I can give a shot if that's helpful.
Chris, any updates with 4.20 kernel?
(In reply to Chris Murphy from comment #11) > The two attachments are based on 4.19.0; Fedora is in the middle of rebasing > to 4.19 so for whatever reason there won't be any 4.19 builds until that > happens. But they have 4.20rc1 built which I can give a shot if that's > helpful. Have you tried to verify with kernel 4.20?
I haven't seen it with 4.19.2 or higher, including 4.20rc5.
Thanks for the feedback Chris. I can close this bug as WORKSFORME.
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.