Avi Kivity wrote:
Anthony Liguori wrote:
Huang Ying wrote:
The related MSRs are emulated. MCE capability is exported via
extension KVM_CAP_MCE and ioctl KVM_X86_GET_MCE_CAP_SUPPORTED. A new
vcpu ioctl command KVM_X86_SETUP_MCE is used to setup MCE emulation
such as the mcg_cap. MCE is injected via vcpu ioctl command
KVM_X86_SET_MCE. Extended machine-check state (MCG_EXT_P) and CMCI are
not simulated.
Maybe I'm missing something, but couldn't this be implemented
entirely within userspace? There's nothing VT/SVM specific about
this. If the issue is setting these MSRs from userspace via
KVM_SET_MSRS isn't enough, perhaps we should add userspace MSR handling.
You also need to inject the MCE.
Regardless of the KVM interface for this, to go upstream to QEMU, this
needs a TCG implementation which means the logic must be duplicated in
userspace. In particular, this is because a user-visible command is
being introduced in the monitor.
So Avi, regardless of what interface is chosen for KVM, can you hold off
applying these patches until there is a TCG implementation? Otherwise,
we'll be playing catchup between kvm-userspace and upstream QEMU forever.
Regards,
Anthony Liguori
--
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
