On Tue, 2015-05-26 at 13:28 +0200, Peter Krempa wrote:
The implementer of the check might remember more reasons for arguing on this, let him shed the light into this discussion. Michal?
Anyway, this is broken all the way because the command also makes sense only with some particular timers and their settins, not with all of them.
In that case it would make more sense to have the interrupt reinjection reset as a separate option or it to be enabled with a flag.
I've talked to Michal and he agrees with the proposed approach.
I don't quite catch your drift here. If you mean that the second condition added should be based on virQEMUCapsGet(), then I agree.
Yes that was my intent. The actual call should be based on the capability. Sanity checks that are arch specific should be present before.
Sure, it works much better that way, thanks for the suggestion :) I've updated the patch accordingly. Cheers. -- Andrea Bolognani Software Engineer - Virtualization Team $ python -c "print('a'.join(['', 'bologn', '@redh', 't.com']))"