From: Gerd H. <kr...@re...> - 2008-05-06 16:00:05
|
Marcelo Tosatti wrote: > F8 host, recent kvm-userspace.git (so with IO thread), recent kvm.git > (plus your patches), haven't tried 2x but I think 4x is not necessary to > reproduce the problem. Ok, see it too. Seem to be actually two (maybe related) problems. First the guest hangs hard after a while, burning 100% CPU time (deadlocked I guess), doesn't respond to sysrq any more. Is there some easy way to get the guest vcpu state then? EIP for starters, preferably with stack trace? The other one is that one ticks slower than the other. I don't see it from start, but after a while it starts happening (unless the guest deadlocks before ...). cheers, Gerd -- http://kraxel.fedorapeople.org/xenner/ |