[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] Thanx for the kqemu patches; some notes about NetBSD/Op
Re: [Qemu-devel] Thanx for the kqemu patches; some notes about NetBSD/OpenBSD guests
Sun, 31 May 2009 16:41:33 +0200
On Sun, May 31, 2009 at 11:29:28AM +0200, Andreas Färber wrote:
> Hi Jürgen,
> Am 30.05.2009 um 22:51 schrieb Juergen Lock:
> > I just committed Jan's kqemu patch series to the FreeBSD kqemu port
> > (yes FreeBSD is another one of those `left behind' OSes where kvm
> > doesn't really work yet...) - and I also committed an old NetBSD/
> > OpenBSD
> > guest kqemu patch,
> > http://lists.gnu.org/archive/html/qemu-devel/2007-11/msg00125.html
> > and then played with a NetBSD guest in qemu 0.10.5 a little, where I
> > found out a few things:
> > 1. A few times I even got NetBSD 5.0/i386 running with -kernel-kqemu
> > now,
> > tho a few other times I also got guest processes segfaulting etc.
> > `Regualr' kqemu seems stable now tho.
> Thanks for the pointer. I've pushed it to a pick-up branch for now to
> facilitate testing. Did you test non-BSD guests, too?
Oh yes I tested Linux too (and FreeBSD), and I tested a Windows 7 beta
guest that I had installed a while ago more out of curiosity than
anything else, and that one seemed to bluescreen a little less than
it used to also... (it usually only bluescreen at boot and sometimes
at shutdown, once it gets past the boot stage it keeps running, altho
slowly. Oh and -kernel-kqemu seems to always make it bluescreen so
only `regular' kqemu is useful for that guest.)
Everything else I tested seemed to still run like it used to, the
Linux guests even _appeared_ to run a little faster than they used to,
possibly because of the added verr/verw and lar/lsl handling in Jan's
> I noticed that Adrian has patches for OpenBSD host support in their
> ports CVS repository, as pointed out by Mikolaj.
> There appears to be some FreeBSD code in-tree though - so, is the
> above patch the only one you've queued, or do you have host patches as
> well that we could push "upstream"?
I have a few more, as you can see on cvsweb:
(patch-tssworkaround is only applied for FreeBSD < 7.1 since later
versions no longer have the issue it works around; look in that patch's
commitlogs if you want links to the gory details...)