David Koudela wrote: > Yes, the dtrace program run on Xorg (see attached dtrace script). > > I have tried analyzing of it in more detail. > The dtrace script was modified to: > - show stack() from do_interrupt of Xorg process > > Your dtrace program is capturing the kernel do_interrupt() routine which is called during real interrupt handling. This is not something caused by the Xorg process.
Cursor movement would be closely associated with interrupts delivered from the USB controller handling the mouse. If the concern is excessive interrupts associated with mouse movement, which gets masked by a reboot, the following info would be useful: # echo "::interrupts" | mdb -k # echo "::prtusb" | mdb -k
