Date: 12 Jul 2001 20:23:38 -0400 From: Greg Troxel <gdt@fnord.ir.bbn.com> To: Nick Sayer <nsayer@quack.kfu.com> Cc: Charles Anderson <caa@columbus.rr.com>, freebsd-emulation@freebsd.org Subject: Re: FreeBSD VMware guest never idles Message-ID: <rmiae291xth.fsf@fnord.ir.bbn.com> In-Reply-To: Nick Sayer's message of "Fri, 09 Mar 2001 13:53:22 -0800" References: <200103091510.f29FADi30516@morpheus.kfu.com> <20010309155114.C63383@midgard.dhs.org> <3AA950D2.5060108@quack.kfu.com>
next in thread | previous in thread | raw e-mail | index | archive | help
(replying to mail from 4 months ago)
Nick Sayer <nsayer@quack.kfu.com> writes:
> Disconnecting the RTC fixed it. I wonder what the mechanism involved is.
I am running vmware2-2.0.3.799_1 under 4.3-stable (early june +
security patches) on a TP600E. I have found that starting up a guest
os (either NetBSD 1.5 or Windows 2000) causes reasonble idle behavior
(i.e., when the load average on the guest os is low, the cpu usage of
the vmware process is also low). After suspending, exiting vmware,
restarting and resuming, I find that the vmware process is running
continuously, even though the NetBSD host is just running top and
showing 0.08 load. All of this is with the RTC connected.
Disconnecting the RTC causes vmware cpu usage to become reasonable,
and reconnecting causes it to become essentially 100%.
So, this seems to have something to do with a resumed VM versus a new
VM. I wonder if there is some RTC 'hardware' state that is not
properly saved across suspend/resume, but is initialized at boot time.
Can anyone explain the consequences of disconnecting RTC? Is having
this 'connected' what makes the "hardware clock" in the VM synced to
the host os operating system time? Should I just run NTP on the guest
instead?
Greg Troxel <gdt@ir.bbn.com>
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-emulation" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?rmiae291xth.fsf>