Daniel Paquet wrote: > Just tried -no-kvmirqchip and it's a no go, and after it I tryed > without taskset and -no-kvmirqchip and it worked... Anyway I will > allways use taskset until now. >
Sorry, hard to piece all the info from the scattered reports. Please report, for each of: - regular - -no-kvm-irqchip - taskset Whether you get - reliable failure - sometimes works, sometimes doesn't - always works so I can try to find a pattern. Also, is tap required to reproduce the problem, or do you occasionally get it with -net user? > On 9/19/07, * Avi Kivity* <[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>> > wrote: > > Daniel Paquet wrote: > > Well well, that did it!!! > > > > For now I even restarted my host machine and the guest os boots > > flawlessly now with taskset! Oh well, I now know a new nice > linux app > > taskset :) > > > > I saw similar problems with some games under windows, that you > had to > > set the affinity to a specific core, if not you were getting those > > kind of small hang, like I was having with kvm and my guest > machines. > > > > And you guessed right I have a dual core cpu. > > > > Anyone have an idea why this is only happening when using a tap > > network interface? I am curious... > > No idea. > > Can you try one more thing? -no-kvm-irqchip, but no taskset. I'm > trying to find out if the smp issue is with the apic code or > elsewhere. > > -- > error compiling committee.c: too many arguments to function > > > > > -- > Ubuntu est un ancien mot africain qui veut dire "Je ne peux pas > configurer Debian" > -- > Daniel Paquet -- Any sufficiently difficult bug is indistinguishable from a feature. ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ kvm-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/kvm-devel
