Bug 514305

Summary: ksystemstats_intel_helper crashes on opening System Monitor
Product: [Frameworks and Libraries] ksystemstats Reporter: Steve Zakulec <spzakulec>
Component: GeneralAssignee: Plasma Bugs List <plasma-bugs-null>
Status: ASSIGNED ---    
Severity: crash CC: ahiemstra, jevgeni.kostenko+kde.org, kdedev, nate, plasma-bugs-null
Priority: HI    
Version First Reported In: 6.5.80   
Target Milestone: ---   
Platform: KDE Linux   
OS: Linux   
Latest Commit: Version Fixed/Implemented In:
Sentry Crash Report:
Attachments: backtrace for ksystemstats_intel_helper crash
backtrace try 2

Description Steve Zakulec 2026-01-08 02:31:02 UTC
SUMMARY
Every time I open System Monitor, there's a popup notification about a ksystemstats_intel_helper crash, and there's consequently no GPU sensor available.

STEPS TO REPRODUCE
1. Open System Monitor on a system with an Intel GPU
2. 
3. 

OBSERVED RESULT
Immediate notification of crash, no GPU usage available on System Monitor Overview page.

EXPECTED RESULT
No crash.

SOFTWARE/OS VERSIONS
Operating System: KDE Linux 2026-01-01
KDE Plasma Version: 6.5.80
KDE Frameworks Version: 6.22.0
Qt Version: 6.10.1
Kernel Version: 6.18.2-zen2-1-zen (64-bit)
Graphics Platform: Wayland
Processors: 4 × Intel® Core™ i3-2370M CPU @ 2.40GHz
Memory: 8 GiB of RAM (7.6 GiB usable)
Graphics Processor: Mesa Intel® HD Graphics 3000
Manufacturer: Dell Inc.
Product Name: Inspiron 3520

ADDITIONAL INFORMATION
Backtrace from most recent crash copied from Crashed Processes Viewer:

           PID: 2856 (ksystemstats_in)
           UID: 1000 (steve)
           GID: 1000 (steve)
        Signal: 8 (FPE)
     Timestamp: Wed 2026-01-07 21:18:08 EST (6min ago)
  Command Line: /usr/lib/ksystemstats_intel_helper 0000:00:02.0
    Executable: /usr/lib/ksystemstats_intel_helper
 Control Group: /user.slice/user-1000.slice/user@1000.service/background.slice/plasma-ksystemstats.service
          Unit: user@1000.service
     User Unit: plasma-ksystemstats.service
         Slice: user-1000.slice
     Owner UID: 1000 (steve)
       Boot ID: 91da770cac894c82b2f6aef56cd7b3ea
    Machine ID: 05ec6b9fbf374094a25f74b8e7c8ffe1
      Hostname: steve-inspiron3520
       Storage: /var/lib/systemd/coredump/core.ksystemstats_in.1000.91da770cac894c82b2f6aef56cd7b3ea.2856.1767838688000000.zst (inaccessible)
       Message: Process 2856 (ksystemstats_in) of user 1000 dumped core.
                
                Stack trace of thread 2856:
                #0  0x000055d37cb71d98 n/a (/usr/lib/ksystemstats_intel_helper + 0x2d98)
                #1  0x00007f16bc427635 n/a (libc.so.6 + 0x27635)
                #2  0x00007f16bc4276e9 __libc_start_main (libc.so.6 + 0x276e9)
                #3  0x000055d37cb725f5 n/a (/usr/lib/ksystemstats_intel_helper + 0x35f5)
                ELF object binary architecture: AMD x86-64
Comment 1 TraceyC 2026-01-08 18:22:06 UTC
Thank you for the bug report. I'm not able to reproduce this on git-master, on a system with an Intel GPU. Unfortunately the backtrace is incomplete and missing debug symbols for the following lines that we need to figure out exactly what's going wrong:

#3  0x000055d37cb725f5 n/a (/usr/lib/ksystemstats_intel_helper + 0x35f5)

Could you please install debug packages and attach a new symbolicated backtrace generated by using `coredumpctl gdb` in a terminal window?

An additional observation is that on my system, while nvtop shows GPU usage when playing videos, the GPU sensor shows nothing. There is no crash, however. On another system with Plasma 6.5.4 and an nvidia GPU, the sensor does show usage.
Comment 2 Steve Zakulec 2026-01-10 16:52:15 UTC
Created attachment 188390 [details]
backtrace for ksystemstats_intel_helper crash
Comment 3 TraceyC 2026-01-13 23:38:45 UTC
Thanks for your reply. Unfortunately, the attachment doesn't contain an actual backtrace. It doesn't look like "bt" was entered.
Can you please create another backtrace and attach that? Thanks!
Comment 4 Steve Zakulec 2026-01-14 01:47:46 UTC
Created attachment 188496 [details]
backtrace try 2

I think this is slightly better- it's again very short, but it does point somewhere now.
Comment 5 Nicolas Fella 2026-01-14 13:54:41 UTC
*** Bug 514580 has been marked as a duplicate of this bug. ***
Comment 6 access 2026-01-14 14:51:41 UTC
This looks like the same SIGFPE (division by zero) issue as Bug 514580:
https://bugs.kde.org/show_bug.cgi?id=514580

That report includes root cause analysis and a minimal patch/workaround.
Comment 7 Nate Graham 2026-01-16 18:18:44 UTC
*** Bug 514580 has been marked as a duplicate of this bug. ***
Comment 8 access 2026-01-17 07:58:33 UTC
I’ve submitted a tested fix as a merge request:
https://invent.kde.org/plasma/ksystemstats/-/merge_requests/120

The patch guards division by zero in ksystemstats_intel_helper (IntelHelper.cpp), which was causing SIGFPE when enginesPerClass[CLASS] == 0.

After applying the fix, plasma-ksystemstats.service no longer crashes and Intel GPU sensors/graphs work correctly again in Plasma System Monitor and widgets.

Duplicate report with additional details and testing:
https://bugs.kde.org/show_bug.cgi?id=514580
Comment 9 Nate Graham 2026-01-17 15:28:09 UTC
Thanks so much!