https://bugs.kde.org/show_bug.cgi?id=516479

--- Comment #4 from [email protected] ---
This was still happening for me on Arch / Plasma 6.6.1 (my specs are in the
duplicate bug https://bugs.kde.org/show_bug.cgi?id=516672). Note this was
happening with the system being awake but the monitors powering down after
inactivity/screenlock.

Interestingly, it seems to have stopped happening after disabling MST on my
displays' onboard  menus (both older Dell UltraSharp LCD panels, both connected
to the GPU via DisplayPort directly, no daisy-chaining so the MST enablement
wasn't doing anything for me).

I noted some oddities in dmesg when the displays were powered back on related
to drm and amdgpu, as well as some powerdevil DDC messages in journalctl
--user. 

Dmesg:
 2221.315037] [drm] DM_MST: stopping TM on aconnector: 000000000b539075 [id:
104]
[ 2222.129177] [drm] DM_MST: starting TM on aconnector: 000000000b539075 [id:
104]
[ 2222.134729] [drm] DM_MST: DP12, 4-lane link detected
[ 6800.207152] [drm] DM_MST: stopping TM on aconnector: 000000000b539075 [id:
104]
[ 6801.029817] [drm] DM_MST: starting TM on aconnector: 000000000b539075 [id:
104]
[ 6801.035371] [drm] DM_MST: DP12, 4-lane link detected
[ 7007.899875] [drm] DM_MST: stopping TM on aconnector: 000000000b539075 [id:
104]
[ 7008.110274] amdgpu 0000:03:00.0: [drm] REG_WAIT timeout 1us * 100 tries -
dcn32_program_compbuf_size line:147

=================
PowerDevil:
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Removing connected
display on bus 11
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Emitting
DDCA_Display_Status_Event[205.012:  DDCA_EVENT_DISPLAY_DISCONNECTED,
card1-DP-5, dref: DDCA_Display_Ref[2], io_path:/dev/i2c-11, ddc working: false]
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Starting 1 callback
threads
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] libddcutil callback
thread 0x7f5774003e40 started
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Started 1 event
callback thread(s)
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 20616] Invoking callback
function 0x7f57a5d4b7f0 for event DDCA_Display_Status_Event[205.012: 
DDCA_EVENT_DISPLAY_DISCONNECTED, card1-DP-5, dref: DDCA_Display_Ref[2],
io_path:/dev/i2c-11, ddc working: false] in this thread [ 20616]
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Udev event detected
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 20616] Callback function
0x7f57a5d4b7f0 for event DDCA_Display_Status_Event[205.012: 
DDCA_EVENT_DISPLAY_DISCONNECTED, card1-DP-5, dref: DDCA_Display_Ref[2],
io_path:/dev/i2c-11, ddc working: false] complete
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Udev event detected
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Udev event detected
Feb 27 09:46:53 cezanne org_kde_powerdevil[19681]: [ 19734] Udev event detected
=================

Obviously whatever causes *that* isn't likely a Plasma issue, but from my naive
understanding it appears to have the effect of hotplugging the display on wake
(by hard-resetting the DP link). After disabling MST, the above messages
haven't recurred and the times are synced across displays when they come up. So
it looks like that DP reset on wake was tripping something up behind the
scenes, at least on my machine?

Hopefully the added context helps!

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to