| Summary: | "Power Management settings could not be loaded" after update to 6.3.1.1 (Assertion failed: "bus_info" in file ddc_packet_io.c at line 266) | ||
|---|---|---|---|
| Product: | [Plasma] plasmashell | Reporter: | JJ <web> |
| Component: | Power management & brightness | Assignee: | Plasma Bugs List <plasma-bugs> |
| Status: | RESOLVED UPSTREAM | ||
| Severity: | normal | CC: | cptncosmo, jared, martin+kde, nate, nicolas.fella, rockowitz, victorsouzaleal, zhaierfang0208 |
| Priority: | NOR | ||
| Version First Reported In: | 6.3.1 | ||
| Target Milestone: | 1.0 | ||
| Platform: | Arch Linux | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ddcutil 2.2.1 | |
| Sentry Crash Report: | |||
| Attachments: |
The error message displayed opening Power Management in settings.
powerdevil systemd status |
||
|
Description
JJ
2025-02-27 22:59:25 UTC
Please show the output of running "systemctl --user status plasma-powerdevil.service" while it's broken Created attachment 178975 [details]
powerdevil systemd status
I'm having the same problem with it, but in my case it's on the Desktop, whenever my screen comes back from sleep, Powerdevil stops working.
Let's let JJ respond before debugging further to make sure that both of you are hitting the same issue (Viktor's issue is an assert in DDC code). (In reply to Nate Graham from comment #3) > Let's let JJ respond before debugging further to make sure that both of you > are hitting the same issue (Viktor's issue is an assert in DDC code). Yes, sorry, it's the first time I've sent a report and I found it a bit confusing, do you think I should do a separate report? If JJ has the same assert in DDC that you have, then there's no need since your issue is the same. If not, then yeah, make a new bug report for your issue. I did some tests and it seems that this is related to a modified EDID so that I can force the use of pixel format RGB. I undid the changes and so far I haven't had the error again. (In reply to Victor from comment #6) > I did some tests and it seems that this is related to a modified EDID so > that I can force the use of pixel format RGB. > > I undid the changes and so far I haven't had the error again. Nah, just happened again :( i'll wait for JJ and then open a new bug report if needed Thanks I rebooted shortly after reporting, and have not run into it again since - I'll update here when it next occurs. Restoring default fonts can avoid crashes I suspect and font settings related I have (finally) re-triggered this bug. It has something to do with external monitors. The output of `systemctl --user status plasma-powerdevil.service` is below:
```bash
○ plasma-powerdevil.service - Powerdevil
Loaded: loaded (/usr/lib/systemd/user/plasma-powerdevil.service; static)
Active: inactive (dead) since Wed 2025-03-12 20:16:08 PDT; 2min 52s ago
Duration: 3d 1h 29min 3.794s
Invocation: e01035822b3544d19a9d159ddab777d0
Process: 1203 ExecStart=/usr/lib/org_kde_powerdevil (code=killed, signal=TERM)
Main PID: 1203 (code=killed, signal=TERM)
Mem peak: 13.3M (swap: 7.1M)
CPU: 1min 1.696s
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: [346881] Emitting DDCA_Display_Status_Event[264544.462: DDCA_EVENT_DISPLAY_CONNECTED, card1-HDMI-A-1, dref: DDCA_Display_Ref[11], io_path:/dev/i2c-1, ddc working: true]
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: [346881] libddcutil callback thread 0x7ab908014220 started
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: [346881] Started 1 event callback thread(s)
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: [346882] (ddc_open_display) Delay before rechecking attribute status: Slept for 1000 millisec
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: [346882] Assertion failed: "bus_info" in file ddc_packet_io.c at line 266
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: QtDBus: cannot relay signals from parent QObject(0x653488af7ab0 "") unless they are emitted in the object's thread QThread(0x653488a4f080 "Qt mainThread"). Current thread is QThread(0x653488bc3118 "").
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: QSocketNotifier: Socket notifiers cannot be enabled or disabled from another thread
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: [346882] Watch thread terminated.
Mar 12 20:16:08 arch org_kde_powerdevil[1203]: QMutex: destroying locked mutex
Mar 12 20:16:08 arch systemd[951]: plasma-powerdevil.service: Consumed 1min 1.696s CPU time, 13.3M memory peak, 7.1M memory swap peak.
```
Yea it looks like we share the same error "assertion failed" this should be fixed in ddcutil 2.2.1; please try again once that reaches your distro! *** Bug 503966 has been marked as a duplicate of this bug. *** *** Bug 504861 has been marked as a duplicate of this bug. *** As described in the discussion of bug 843861, libddcutil macros wrapping an assertion did not call call __assert_fail(). Apparently as a result, powerdevil failed with status success and did not restart. This has been fixed in libddcutil branch 2.2.1-dev. Also, the function in question has been modified to return a status code instead of asserting failure in the situation in question, so the (ultimately) invoking API function should complete with an error status indicating the display in question has been disconnected. |