Vkernel issues

Matthew Dillon dillon at apollo.backplane.com
Sun Jul 22 13:06:16 PDT 2007

:Hi all,
:I have just updated a dual-core Opteron machine to DragonFly-1.9 to play
:with the new SMP-improved virtual kernels.
:First, I tried to map the virtual CPUs to the real ones and failed:
:./kernel.debug -m 128m -r /var/vkernel/rootimg.01 -I auto:bridge0 -n 2 -l map
:Bad target CPU number at '0�&(P�&(��&(p�&(�C$(�[$]$((]$(�_$(�b$(�c$(��$(�$(!�$(.x$(�x$(�B$('

    The vkernel is not parsing the option properly.  I'll commit a fix in
    a second.

:The second issue is a minor one.
:Virtual kernel options: -I auto:bridge0 -n 2 -l any
:top capture on a virtual machine :
:load averages:  0.94,  0.26,  0.09                     up 0+00:03:36  17:53:57
:10 processes:  10 running
:CPU0 states:  0.0% user,  0.0% nice,  100% system,  0.0% interrupt,  0.0% idle
:CPU1 states:  0.0% user,  0.0% nice,  9.7% system,  0.0% interrupt, 90.3% idle
:The first CPU is almost always shown with a 100% system time.
:This is also true with a UP vkernel.
:Francois Tigeot

    Yah, this is a known issue.  The vkernel's cpu statistics are having
    trouble figuring out what context the vkernel is in.  The cpus are not
    actually 100% system and 90% idle.

					Matthew Dillon 
					<dillon at backplane.com>

More information about the Kernel mailing list