Version: (using KDE Devel) Installed from: Compiled sources If a driver misbehaves, the cpu load average can be very high but with no process using the cpu usage. That can be very confusing, and not at all obvious on how to track it down. The usual reason is that there is a misbehaving module. The first step is to check /proc/interrupts On a well behaved system, it will look like: 0: 6665963 6677722 IO-APIC-edge timer 1: 12838 13062 IO-APIC-edge i8042 7: 0 0 IO-APIC-edge parport0 8: 0 0 IO-APIC-edge rtc 9: 0 0 IO-APIC-fasteoi acpi 12: 1115 1106 IO-APIC-edge i8042 16: 8384933 8369671 IO-APIC-fasteoi uhci_hcd:usb4, eth1, HDA Intel, i915@pci:0000:00:02.0 18: 429823 429320 IO-APIC-fasteoi uhci_hcd:usb3, libata 19: 1266898 1270584 IO-APIC-fasteoi uhci_hcd:usb2, libata, wifi0 23: 1 2 IO-APIC-fasteoi uhci_hcd:usb1, ehci_hcd:usb5 509: 19521 19499 PCI-MSI-edge eth2 NMI: 0 0 LOC: 13342951 13342933 ERR: 0 The number is the number of interrupts. Notice that the interrupts are split evenly between the 2 processors, and the numbers are fairly low. This system has been up for 14 hours. A 'bad' line would look like: 169: 294912182 612411557 IO-APIC-level ohci_hcd, ohci_hcd This information should be intelligently checked and shown to the user somehow.
ksysguard is no longer maintained, in Plasma 6 there is the Plasma system monitor for this task. If your wish is still valid for the Plasma 6 replacement, please re-open and we can move this bug to the new product, thanks!