On Sun, Aug 07, 2011, Joerg Roedel wrote about "Re: [PATCH 3/3] Fix TSC MSR read in nested SVM": > > Joerg, can you review and ack Nadav's SVM patch? TIA > > Tested-by: Joerg Roedel <joerg.roedel@xxxxxxx> > Acked-by: Joerg Roedel <joerg.roedel@xxxxxxx> > > Reviewed and tested (didn't apply cleanly, but was easy to fix that up). > Looks all good so far. Hi guys, I'm a bit confused how we want to proceed from here. The patches which I sent appear to fix the original bug (as confirmed by the two people who reported it), but Zachary warned that it would break migration of nested SVM while L2 is running. I asked whether migration works at all while L2 is running (without exiting to L1 first) - and Marcelo suggested that it doesn't. If the problem Zachary pointed to is considered serious, I proposed a second option - to leave the code to *wrongly* return the L1 TSC MSR always, and check (and warn) in situations where this value is wrongly returned to the guest, but this will leave qemu to always read the TSC MSR from L1, even when L2 is running. While I proposed this as a second option, I don't think I can recommend it. So what's the verdict? :-) Thanks, Nadav. -- Nadav Har'El | Wednesday, Aug 10 2011, 10 Av 5771 nyh@xxxxxxxxxxxxxxxxxxx |----------------------------------------- Phone +972-523-790466, ICQ 13349191 |It's fortunate I have bad luck - without http://nadav.harel.org.il |it I would have no luck at all! -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
