Performance Issues, MacFUSE related?

[non-troppo]

On my friends Macbook with latest V3.0 build (he has never installed MacFUSE), there are occasional cases where Parallels CPU goes up, windowserver CPU also increases -- in these cases the mouse goes crazy, jumping all over the place. Today there was a warning dialog stating that MacFUSE could not connect, and he managed to get this sample while Parallels was not responding:

Analysis of sampling pid 666 every 10.000000 milliseconds
Call graph:
296 Thread_0f07
296 0x66b9
296 0x6792
296 0x41dc9
296 0x388d5d
296 0x5ab264
296 0x5ab14b
296 0x5fec31
296 0x40755a
296 SendEventToEventTarget
296 SendEventToEventTargetInternal(OpaqueEventRef*, OpaqueEventTargetRef*, HandlerCallRec*)
296 DispatchEventToHandlers(EventTargetRec*, OpaqueEventRef*, HandlerCallRec*)
296 ToolboxEventDispatcherHandler(OpaqueEventHandlerCallRef*, OpaqueEventRef*, void*)
296 SendEventToEventTarget
296 SendEventToEventTargetInternal(OpaqueEventRef*, OpaqueEventTargetRef*, HandlerCallRec*)
296 DispatchEventToHandlers(EventTargetRec*, OpaqueEventRef*, HandlerCallRec*)
296 0x40eb4e
296 0x375ca1
296 0x370915
296 0x382684
296 0x386536
296 0x3aa1c1
296 0x3856dd
296 0x3854ef
296 0x583ebb
296 0x3796d7
296 0x3542d3
296 0x12d5f3
296 0x32a58d
296 0x32a51e
296 0x550a56
296 0x54d735
296 0x54d6bd
296 unmount
296 unmount
296 Thread_1003
296 _pthread_body
296 CAPThread::Entry(CAPThread*)
296 HALRunLoop::OwnThread(void*)
296 CFRunLoopRunInMode
296 CFRunLoopRunSpecific
296 mach_msg_trap
296 mach_msg_trap
296 Thread_1103
296 _pthread_body
296 0x3b6836
296 0x45508c
253 0x45465d
253 semaphore_timedwait_signal_trap
253 semaphore_timedwait_signal_trap
43 ioctl
43 ioctl
296 Thread_1203
296 _pthread_body
296 0x3b6836
296 0x43b9e2
296 0x64bb3c
295 0x648c0a
295 usleep
295 mach_wait_until
295 mach_wait_until
1 0x648c20
1 0x648c20
296 Thread_1303
296 _pthread_body
296 0x3b6836
296 0x64be68
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1403
296 _pthread_body
296 0x3b6836
296 0x64be68
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1503
296 _pthread_body
296 0x3b6836
296 0x64be68
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1603
296 _pthread_body
296 0x65f6cc
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1703
296 _pthread_body
296 0x3b6836
296 0x64be68
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1803
296 _pthread_body
296 0x725401
296 QList<TGDev>::append(TGDev const&)
296 0x659eb4
296 0x7251cc
296 semaphore_timedwait_trap
296 semaphore_timedwait_trap
296 Thread_1903
296 _pthread_body
296 0x3b6836
296 0x64be68
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1a03
296 _pthread_body
296 0x725401
296 QList<TGDev>::append(TGDev const&)
296 0x659eb4
296 0x7251cc
296 semaphore_timedwait_trap
296 semaphore_timedwait_trap
296 Thread_1b03
296 _pthread_body
296 0x3b6836
296 0x64be68
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1c03
296 _pthread_body
296 0x725401
296 QList<TGDev>::append(TGDev const&)
296 0x659eb4
296 0x7251cc
296 semaphore_timedwait_trap
296 semaphore_timedwait_trap
296 Thread_1d03
296 _pthread_body
296 0x3b6836
296 0x64be68
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap
296 Thread_1e03
296 _pthread_body
296 0x3b6836
296 0x43ac32
296 CFRunLoopRun
296 CFRunLoopRunSpecific
296 mach_msg_trap
296 mach_msg_trap
296 Thread_1f03
296 _pthread_body
296 0x3b6836
296 0x43b905
296 usleep
295 mach_wait_until
295 mach_wait_until
1 nanosleep
1 nanosleep
296 Thread_2003
296 _pthread_body
296 0x45a2e1
296 0x54dd0b
296 0x54d22c
296 0x6b2fa6
296 0x52e8ea
296 semaphore_wait_signal_trap
296 semaphore_wait_signal_trap

Total number in stack (recursive counted multiple, when >=5):
17 _pthread_body
11 0x3b6836
9 semaphore_wait_signal_trap
7 0x64be68

Sort by top of stack, same collapsed (when >= 5):
semaphore_wait_signal_trap 2664
semaphore_timedwait_trap 888
mach_msg_trap 592
mach_wait_until 590
unmount 296
semaphore_timedwait_signal_trap 253
ioctl 43
Sample analysis of process 666 written to file /dev/stdout
Sampling process 666 each 10 msecs 300 times